OpenAI o4-mini
Last reviewed
May 17, 2026
Sources
25 citations
Review status
Source-backed
Revision
v2 · 6,199 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
May 17, 2026
Sources
25 citations
Review status
Source-backed
Revision
v2 · 6,199 words
Add missing citations, update stale details, or suggest a clearer explanation.
OpenAI o4-mini is a compact reasoning model developed by OpenAI, released on April 16, 2025. It belongs to the o-series line of models that use test-time compute to deliberate before answering, and it succeeds o3-mini as OpenAI's small, cost-efficient reasoning option. Unlike its predecessors in the mini tier, o4-mini ships with full tool access built in: it can browse the web, execute Python, process and generate images, and chain those tools together in a single reasoning trace. It also introduces native visual reasoning, meaning the model can incorporate images into its chain of thought and apply transformations such as cropping or zooming mid-reasoning. On AIME 2025 without external tools, o4-mini scores 92.7%, outperforming OpenAI o3 (88.9%) and its predecessor o3-mini (86.5%). With a Python interpreter attached, it reaches 99.5% on the same benchmark.
Notably, OpenAI did not release a standalone "o4" model alongside o4-mini. The full-size fourth-generation reasoning capability was rolled into later products, leaving o4-mini as the sole representative of the o4 generation at launch. o4-mini was retired from ChatGPT on February 13, 2026, with API endpoints removed on February 16, 2026. The recommended successors became GPT-5 mini and GPT-5.1 mini for general use, and the o3-mini snapshot for the closest functional drop-in on reasoning-heavy workloads. Despite its short ChatGPT lifespan, o4-mini was one of the most heavily adopted reasoning models of the 2025 cycle thanks to inclusion as the default free-tier reasoning option in GitHub Copilot, Cursor, and Microsoft Azure AI Foundry.
OpenAI introduced the o-series in September 2024 with OpenAI o1 and o1-mini. These models departed from the pattern of scaling parameters alone. Instead, they used reinforcement learning to teach the model to generate long internal reasoning chains before producing an answer, a technique sometimes called chain-of-thought at inference time or test-time compute scaling. The tradeoff was deliberate: o1 was slower and more expensive than GPT-4o, but it was substantially more reliable on complex math, science, and coding tasks.
The naming convention has caused repeated confusion. OpenAI skipped "o2" entirely, reportedly to avoid trademark conflicts with the British mobile carrier O2. The progression went o1 (September 2024), o3 and o3-mini (January 2025), and then o4-mini (April 2025). There was no o4 full model at launch. This means that at the time of o4-mini's release, OpenAI's lineup had o3 as its most powerful reasoning model and o4-mini as the efficient complement, even though the latter carried a higher generation number. The confusion is real: o3 outperforms o4-mini on several benchmarks, despite the version numbers implying the opposite.
Each generation of the o-series added capability. o1 could reason but could not use tools. o3 and o3-mini could do some tool-assisted tasks. o4-mini was the first mini model in the series to support full tool access natively, including the ability to incorporate images into its internal reasoning rather than merely treating them as static input.
The underlying technical approach involves reinforcement learning on reasoning chains. Rather than training the model to predict the next token in a fixed way, the o-series training process rewards the model for reaching correct answers through reasoning steps, even when those steps take unexpected paths. The result is a model that can discover novel solution strategies for problems it has not seen before, rather than pattern-matching to familiar solution templates. This is part of why o-series models perform particularly well on competition mathematics, where novel approaches are required, versus MMLU-style factual recall, where standard language model training excels.
OpenAI trained o4-mini with the same reinforcement-learning pipeline as o3 but at a smaller parameter scale, mirroring the relationship between o1 and o3-mini. Both o3 and o4-mini share a knowledge cutoff of May 31, 2024.
OpenAI announced o3 and o4-mini together on April 16, 2025. Both models were made available immediately to ChatGPT Plus, Pro, and Team subscribers, as well as through the API via the Chat Completions endpoint and the newer Responses API.
On April 24, 2025, OpenAI extended o4-mini access to all ChatGPT users, including the free tier. Free users could access the model through a "Think" mode toggle for more demanding queries. The broader rollout also made o4-mini available through the Chat Completions and Responses APIs without special access requirements.
On the same day as the general release, OpenAI made o4-mini available on Microsoft Azure AI Foundry and as a public preview model in GitHub Copilot on all paid Copilot plans. GitHub's integration allowed developers to query the model from within their coding environments alongside o3.
On June 26, 2025, OpenAI announced the launch of o4-mini-deep-research in the Responses API, with developer access opening the next day. The initial snapshot was o4-mini-deep-research-2025-06-26, and a second snapshot, o4-mini-deep-research-2025-10-10, was released on October 10, 2025, with improvements to multi-hop citation handling and source de-duplication. The variant was fine-tuned for multi-step research workflows, capable of decomposing a question into sub-queries, browsing multiple sources sequentially, and synthesizing a structured report. It is priced at $2 per million input tokens and $8 per million output tokens.
On January 29, 2026, OpenAI announced the retirement of o4-mini from ChatGPT, effective February 13, 2026. The announcement bundled o4-mini's retirement with GPT-4o, GPT-4.1, and GPT-4.1 mini, framing it as a lineup consolidation following the launch of GPT-5 mini and later GPT-5.1 mini. After February 13, 2026, ChatGPT users could no longer select o4-mini or o4-mini-high, although prior chat history remained accessible. API calls to o4-mini and o4-mini-2025-04-16 began returning deprecation responses on February 16, 2026. Recommended successors included GPT-5 mini and GPT-5.1 mini for general use, the o3-mini snapshot for the closest functional drop-in, and gpt-5-codex for coding-centric tasks. The o4-mini-deep-research snapshots remained available beyond the base model retirement.
Three variants of o4-mini were released or became available through API endpoints:
| Variant | Model ID | Tier | Pricing (input / output, per 1M) | Notes |
|---|---|---|---|---|
| o4-mini | o4-mini-2025-04-16 | Free, Plus, Pro, Team, Enterprise, API | $1.10 / $4.40 | Default snapshot; standard reasoning effort |
| o4-mini-high | Same underlying weights, reasoning_effort=high | Plus, Pro, Team, Enterprise, API | $1.10 / $4.40 | Slower; deeper chain-of-thought traces |
| o4-mini-deep-research | o4-mini-deep-research-2025-06-26, o4-mini-deep-research-2025-10-10 | API (Responses) | $2.00 / $8.00 | Agentic research; background mode |
o4-mini is the standard version. It is optimized for speed and throughput. The model ID used in API calls is o4-mini-2025-04-16, and o4-mini resolves to this snapshot. It supports up to 200,000 input tokens and up to 100,000 output tokens.
o4-mini-high applies increased reasoning effort at inference time, producing more thorough chain-of-thought traces before outputting an answer. It is slower than the base o4-mini and was initially restricted to paid ChatGPT subscribers. In the API, o4-mini-high uses the same underlying weights but with a different inference configuration. Both variants carry identical pricing. In Cursor, o4-mini-high was exposed to free-tier users by default, making the high-effort variant unusually accessible compared to its restricted availability elsewhere.
o4-mini-deep-research is a separately fine-tuned variant designed for agentic research tasks. It orchestrates web searches, synthesizes findings from multiple sources, and produces structured reports with inline citations. It is only available through the Responses API and is intended to run in "background" mode; a typical call takes five to twenty minutes to complete.
For API users, the reasoning_effort (Chat Completions API) and reasoning.effort (Responses API) parameter accepts values of low, medium, or high, controlling how many internal reasoning tokens the model is allowed to generate. OpenAI advises starting with medium, dropping to low for high-throughput latency-sensitive flows, and escalating to high only when complex multi-step problems demand it. The max_output_tokens parameter caps total generation including reasoning, and a response that exhausts the budget mid-reasoning returns with status incomplete.
o4-mini uses reinforcement-learning-trained chain-of-thought reasoning. Before producing its final answer, the model works through a reasoning trace that may span hundreds or thousands of tokens. This trace is not shown to users by default, though the Responses API exposes reasoning token usage separately from output tokens. The reasoning is what enables o4-mini to handle multi-step problems that would trip up models that answer in a single pass.
OpenAI describes o4-mini as optimized for "fast, cost-efficient reasoning with exceptionally efficient performance in coding and visual tasks." The emphasis on speed and cost-efficiency distinguishes it from o3, which is positioned as the higher-accuracy option when cost is less of a concern.
Previous mini models in the o-series, including o1-mini and o3-mini, had limited or no tool access in ChatGPT. o4-mini changed this. In ChatGPT, o4-mini can autonomously invoke four categories of tools:
Critically, these tools can be chained. A single response can involve browsing for information, writing code to process it, and generating an image to illustrate the result, all within one continuous reasoning trace. This agentic capability was a significant step beyond previous mini models.
In the OpenAI API, tool use is available via function calling, which was already supported before o4-mini. The model handles structured outputs and streaming. Fine-tuning and embeddings are not available for o4-mini. Parallel tool calling is supported, meaning the model can issue multiple tool calls in a single reasoning step rather than serializing them. This is particularly useful for batch operations such as fetching multiple URLs or running independent code snippets concurrently.
The tool chaining capability is more than a convenience feature. It changes how the model approaches problems. When browsing is available, the model can ground its answers in current information rather than training data. When the Python interpreter is available, it can offload computation rather than performing arithmetic in text space (where errors accumulate across steps). Independent testers found that some tasks that o4-mini completed incorrectly without tools were resolved correctly once tools were enabled, because the model recognized where its reasoning was uncertain and sought external verification.
OpenAI's BrowseComp benchmark, which tests the ability to locate hard-to-find information on the web through iterative browsing, showed o-series models with tool access substantially outperforming base models: o3 with Python and web search scored 49.7%, compared to 1.9% for GPT-4o with basic browsing. This gap reflects how much tool-augmented reasoning differs from simple retrieval.
o4-mini accepts both text and image inputs. It is classified as a vision-language model. Image inputs can be provided inline (base64-encoded) or as URLs in API requests.
The model supports the full range of standard vision tasks: reading text in images, interpreting charts and graphs, identifying objects, and reasoning about spatial relationships. Where o4-mini differs from earlier vision models is in how it uses those images during reasoning.
OpenAI introduced a capability called "thinking with images" alongside o3 and o4-mini, describing it as a new approach to visual problem-solving. In conventional vision models, an image is processed at input and the model's internal computation proceeds in text token space from there. In o4-mini, the model can interact with images as part of its reasoning chain.
Concretely, this means the model can apply image transformations mid-reasoning. If a whiteboard photo is blurry, the model can zoom in on a specific region. If a diagram is rotated, the model can correct the orientation before analyzing it. If a chart has a region of interest, the model can crop to it. These operations happen as tool calls within the chain of thought, not as preprocessing steps applied before the model sees the input.
OpenAI demonstrated several practical scenarios. A handwritten equation on a whiteboard, even if photographed at an angle with poor lighting, can be read and solved. A low-resolution scan of a printed table can be processed by zooming into individual cells. A hand-drawn sketch of a circuit diagram can be analyzed even if the sketch is approximate.
This matters for real-world inputs, which are often imperfect. Previous vision models handled ideal inputs well but degraded on noisy, rotated, or partial images. The reasoning-with-image-tools approach gives o4-mini more flexibility with degraded inputs.
On visual benchmarks, o4-mini scores 81.6% on MMMU (Massive Multidisciplinary Multimodal Understanding) and 84.3% on MathVista, a benchmark for mathematical visual reasoning. These scores trail o3 but represent substantial improvements over o1 (77.6% MMMU, 71.8% MathVista).
OpenAI also reported improvements on VLMs are Blind, a benchmark specifically designed to test basic visual perception tasks that were found to trip up earlier multimodal models, and on V*, a visual search benchmark that requires locating specific elements within complex images. The common theme is that incorporating images into the reasoning process, rather than simply at input, lets the model catch and correct perceptual errors before committing to an answer.
For developers, the practical implication is that o4-mini handles image inputs that previous models would have struggled with. A photo of a handwritten problem taken in poor lighting, a chart from a scan with compression artifacts, or a whiteboard photographed at an angle are all cases where the model's ability to apply corrections mid-reasoning provides a meaningful accuracy advantage.
The following table summarizes o4-mini's performance on key benchmarks at the time of release, compared to o3 and o3-mini. All scores without the "(tools)" annotation are zero-shot without external tool access.
| Benchmark | o4-mini | o3 | o3-mini |
|---|---|---|---|
| AIME 2024 (math olympiad) | 93.4% | 91.6% | 87.3% |
| AIME 2025 (math olympiad) | 92.7% | 88.9% | 86.5% |
| AIME 2025 (with Python) | 99.5% | 98.4% | n/a |
| Codeforces (competitive programming, ELO) | 2719 | 2706 | 2073 |
| SWE-bench Verified (software engineering) | 68.1% | 69.1% | 49.3% |
| GPQA Diamond (PhD-level science) | 81.4% | 83.3% | 79.7% |
| MMMU (multimodal) | 81.6% | 82.9% | n/a |
| MathVista (visual math) | 84.3% | 86.8% | n/a |
| CharXiv (chart reasoning) | 72.0% | 78.6% | n/a |
| Humanity's Last Exam (with tools) | 17.7% | 24.9% | n/a |
| BrowseComp (agentic browsing, with tools) | 49.7%* | 51.5%* | n/a |
| Aider Polyglot (whole, o4-mini-high) | 68.9% | 79.6% | n/a |
| Aider Polyglot (diff, o4-mini-high) | 58.2% | 72.9% | n/a |
| Tau-Bench Telecom (pass^1) | ~50% | ~52% | n/a |
*BrowseComp scores reported for o-series models with browsing and Python tools enabled.
Several patterns emerge from these results. o4-mini leads on AIME 2024 and matches o3 closely on AIME 2025, establishing it as the strongest math benchmark performer in the mini tier and competitive with the full-size o3. On coding benchmarks (Codeforces and SWE-bench), the gap between o4-mini and o3 is narrow. On tests requiring deeper multi-step reasoning across diverse domains (Humanity's Last Exam, CharXiv, Aider Polyglot), o3 holds a more meaningful lead. This pattern aligns with OpenAI's positioning: o4-mini is the better choice for math and coding at lower cost; o3 is better for the most complex cross-domain reasoning.
The Python tool result on AIME is particularly notable. 99.5% is close to a perfect score on a test that challenges the world's best high school mathematicians. The model uses Python to verify intermediate computations, which eliminates arithmetic errors that plague purely generative reasoning.
For context, AIME problems require students to find an integer answer between 0 and 999, with no partial credit. Human competitors at the AIME level are already exceptional students; perfect scores are rare even at international olympiad level. The 99.5% pass@1 score represents near-complete reliability on a problem set that most professionals with mathematical training could not complete in the allotted time.
The Codeforces ELO of 2719 provides another reference point. Codeforces is a competitive programming platform where ELO reflects the strength of a programmer relative to the global competition community. An ELO above 2700 corresponds roughly to the level of a Grandmaster competitor, placing o4-mini among the top competitive programmers globally as measured by this metric.
On Aider Polyglot, a benchmark covering 225 of the hardest Exercism problems across C++, Go, Java, JavaScript, Python, and Rust, o4-mini-high posted 68.9% in whole-file edit mode and 58.2% in diff mode. Both scores trail o3 but place o4-mini-high ahead of every contemporaneous Anthropic mini model and Google's Gemini 2.5 Flash. On the τ²-Bench Telecom evaluation, which simulates a customer-service agent navigating policy documents and API calls, o4-mini achieved a pass^1 of roughly 50%, on par with Claude 3.7 Sonnet and GPT-4.1 mini. Once GPT-5 launched in August 2025, the GPT-5 mini variant overtook o4-mini on most coding and agent benchmarks, although o4-mini retained an advantage on raw AIME 2025 pass-rate without tools.
OpenAI priced o4-mini at the same level as o3-mini at launch, positioning it as the affordable reasoning option in the post-o3 lineup.
| Model | Input (per 1M tokens) | Output (per 1M tokens) | Cached input |
|---|---|---|---|
| o4-mini | $1.10 | $4.40 | $0.275 |
| o4-mini-high | $1.10 | $4.40 | $0.275 |
| o4-mini-deep-research | $2.00 | $8.00 | $0.50 |
| o3 | $10.00 | $40.00 | $2.50 |
| o3-mini | $1.10 | $4.40 | $0.55 |
| o1 | $15.00 | $60.00 | $7.50 |
| GPT-4o | $2.50 | $10.00 | $1.25 |
| GPT-5 mini (Aug 2025) | $0.25 | $2.00 | $0.025 |
Compared to o3, o4-mini costs approximately 89% less per token while delivering competitive or superior performance on several benchmarks. OpenAI described this as roughly a 10x cost reduction relative to o3 for similar workloads. For high-volume applications where reasoning quality matters but the highest possible accuracy is not required, o4-mini offered a substantially more practical price point.
The arrival of GPT-5 mini in August 2025 reshaped the small-reasoning tier. GPT-5 mini undercut o4-mini by roughly 77% on input pricing and 55% on output pricing while extending the context window to 400,000 tokens. This drove most new development away from o4-mini after August 2025, although existing integrations often remained on o4-mini until the formal February 2026 retirement.
Cached input tokens are priced at $0.275 per million, a 75% discount on first-touch input pricing. Batch API requests receive a 50% discount on both input and output tokens, and the Flex pricing tier introduced mid-2025 provided an additional 50% discount in exchange for higher latency tolerance, particularly popular for o4-mini deep-research runs.
Reasoning tokens, which are the internal chain-of-thought tokens generated but not surfaced in the output, count toward token usage. The Responses API exposes separate reasoning_tokens and output_tokens counts, which is useful for understanding actual cost per request on complex tasks.
OpenAI released o3 and o4-mini on the same day, positioning them as complementary rather than directly competing. The following table compares the two models across key dimensions.
| Dimension | o4-mini | o3 |
|---|---|---|
| Input price (per 1M tokens) | $1.10 | $10.00 |
| Output price (per 1M tokens) | $4.40 | $40.00 |
| Context window | 200,000 tokens | 200,000 tokens |
| Max output tokens | 100,000 | 100,000 |
| AIME 2025 (no tools) | 92.7% | 88.9% |
| AIME 2025 (with Python) | 99.5% | 98.4% |
| SWE-bench Verified | 68.1% | 69.1% |
| Codeforces ELO | 2719 | 2706 |
| GPQA Diamond | 81.4% | 83.3% |
| CharXiv | 72.0% | 78.6% |
| Humanity's Last Exam (with tools) | 17.7% | 24.9% |
| Aider Polyglot (whole) | 68.9% | 79.6% |
| Tool use (ChatGPT) | Full | Full |
| Vision input | Yes | Yes |
| Knowledge cutoff | May 31, 2024 | May 31, 2024 |
| Rate limits | Higher | Lower |
| Best for | High-volume math/coding | Complex multi-domain reasoning |
The rate limit difference is meaningful for production use. OpenAI provided higher tokens-per-minute limits for o4-mini, making it more suitable for applications that need to process many requests in parallel. o3 was limited to lower throughput, reflecting the greater compute cost per token.
For most math and coding tasks, o4-mini and o3 perform within a few percentage points of each other. The larger gap appears on tasks requiring sustained reasoning across diverse knowledge domains, like Humanity's Last Exam, where o3's advantage (24.9% vs. 17.7%) is more substantial. The practical recommendation from OpenAI and independent reviewers is to use o4-mini as the default choice and reserve o3 for tasks where that extra performance margin justifies the cost increase.
GPT-5 mini launched in August 2025 as part of the consolidated GPT-5 lineup and became the most direct competitor to o4-mini in the affordable-reasoning tier. The two models target overlapping workloads but reflect different architectural lineages: o4-mini descends from the o-series reasoning-specific training pipeline, while GPT-5 mini is a small variant of the unified GPT-5 architecture that merged reasoning and conversational behaviors.
| Dimension | o4-mini | GPT-5 mini |
|---|---|---|
| Release date | April 16, 2025 | August 7, 2025 |
| Input price (per 1M tokens) | $1.10 | $0.25 |
| Output price (per 1M tokens) | $4.40 | $2.00 |
| Context window | 200,000 tokens | 400,000 tokens |
| Max output tokens | 100,000 | 128,000 |
| Knowledge cutoff | May 31, 2024 | August 31, 2025 |
| SWE-bench Verified | 68.1% | ~71% |
| Reasoning effort tuning | low/medium/high | minimal/low/medium/high |
| Native ChatGPT availability | Retired Feb 13, 2026 | Available |
| Tool chaining | Yes | Yes |
| Multimodal input | Text + image | Text + image |
The pricing gap is dramatic: GPT-5 mini is approximately 4.4x cheaper on input tokens and 2.2x cheaper on output tokens, with roughly double the context window. GPT-5 mini also added a minimal reasoning effort tier below o4-mini's low, allowing developers to disable extended chain-of-thought entirely. On most coding and agentic benchmarks, GPT-5 mini matches or slightly exceeds o4-mini-high while running noticeably faster.
Where o4-mini retained a measurable edge was on pure-reasoning math without tools (AIME 2025 at 92.7%) and on tasks that benefit from o-series-specific reinforcement learning. By the time of o4-mini's ChatGPT retirement in February 2026, OpenAI's migration guidance recommended GPT-5 mini for most o4-mini workloads, with GPT-5 Codex for coding agents and the o3-mini snapshot as a fallback for workloads sensitive to behavioral drift.
o4-mini's combination of strong coding benchmarks and tool access makes it well-suited for software development workflows. In coding assistants like GitHub Copilot, it can generate, explain, and debug code across common programming languages. Its ability to execute Python during reasoning means it can verify that generated code runs correctly before outputting it, rather than producing code that looks plausible but fails at runtime.
On SWE-bench Verified, which measures the ability to resolve real GitHub issues from open-source repositories, o4-mini scores 68.1%. This puts it ahead of Claude 3.7 Sonnet (62.3%) and well ahead of o3-mini (49.3%). For automated code review, test generation, and refactoring tasks, these benchmark scores translate to meaningful reliability improvements over previous mini models.
In Cursor, o4-mini was added on April 16, 2025 and exposed to free-tier users as the o4-mini-high variant by default. Users reported response-time improvements of roughly 15 to 20 percent relative to o3-mini. Free access in Cursor combined with paid access in Copilot gave o4-mini a broader developer audience than most mini reasoning models, and many open-source projects adopted o4-mini as their default automated code-review model through the second half of 2025.
Math is o4-mini's strongest domain. Its near-perfect AIME scores with Python access position it at the frontier of automated mathematical problem-solving. This has practical applications beyond olympiad problems: financial modeling, scientific computing, statistical analysis, and any domain where multi-step arithmetic or algebraic reasoning is required.
The Python tool integration is the key mechanism here. When a problem requires computing a large sum, factoring a polynomial, or running a numerical simulation, the model can delegate to the interpreter rather than attempting the computation in its reasoning trace. The result is fewer arithmetic errors and more reliable answers on quantitative tasks.
In scientific contexts, o4-mini's visual reasoning and high GPQA Diamond score (81.4%) make it useful for interpreting figures, analyzing experimental data, and working through quantitative problems. The deep-research variant extends this to literature review workflows, where the model can autonomously search academic databases, extract relevant findings, and produce structured summaries.
For researchers running analyses across large corpora, o4-mini's higher rate limits and lower cost (compared to o3) make it practical for batch processing. Applying reasoning at scale to hundreds of papers or data files is economically feasible in a way that full-size o3 is not.
The ability to chain tools within a single reasoning trace opens o4-mini to agentic applications. A task might require the model to search for recent data, parse a PDF, run a calculation, generate a chart, and compose a summary, with each step informing the next. o4-mini can execute this sequence autonomously rather than requiring a human to route between tools.
METR, an AI safety organization, evaluated both o3 and o4-mini on autonomous task completion. Their assessment defined a "time horizon score" as the task duration at which a model completes with 50% reliability. o4-mini achieved approximately a 1 hour 15 minute horizon, meaning it could reliably complete agentic tasks that a human would finish in about that time. This is 1.5x the horizon of Claude 3.7 Sonnet in the same evaluation.
In production deployments, o4-mini was widely used as the reasoning engine for customer-support agents, internal knowledge assistants, and developer-tooling pipelines. Its 50% pass^1 on τ²-Bench Telecom was high enough that several telecom and utility companies deployed o4-mini for tier-one support triage, escalating complex flows to o3. Tool-chaining behaviors also made o4-mini a common choice for browser-automation tasks.
Students and educators have applied o4-mini to tutoring, problem-set generation, and explanation tasks. Its chain-of-thought reasoning, when surfaced through the API, shows the steps taken to reach an answer, which serves as a worked example. The visual reasoning capabilities extend this to diagrams and handwritten notes, both common in educational contexts.
o4-mini can combine browsing, Python, and visual reasoning in data analysis workflows. A user can upload a spreadsheet or image of a chart and ask the model to analyze trends, compute statistics, or generate a Python script that processes the data in a specified way. For business intelligence applications, this reduces the gap between asking a question about data and getting an analytical answer: the model handles the intermediate steps of writing code, running it, and interpreting results, rather than simply suggesting what code to write.
The 200,000-token context window supports analysis of large documents. In one demonstration cited at the time of release, o4-mini processed the full 117,649-token Stanford AI Index Report within a single context window and answered detailed questions about its findings in about nine seconds.
Initial reception was largely positive, particularly from developers who had been waiting for a mini model with full tool access. The combination of near-o3 performance on math and coding with a 10x cost reduction made o4-mini immediately practical for a range of applications that o3's price had placed out of reach.
GitHub's simultaneous launch of o4-mini in Copilot on April 16, 2025, brought the model to a large developer audience on the same day as the API launch. Coverage from TechCrunch, The Verge, and VentureBeat highlighted the "thinking with images" capability as the most novel feature, framing it as a meaningful departure from how earlier vision models handled visual inputs.
However, a significant criticism emerged within days of the release. On April 18, 2025, TechCrunch reported that OpenAI's internal PersonQA benchmark showed o4-mini hallucinating at a rate of 48%, substantially higher than the 16% rate for o1 and the 14.8% rate for o3-mini. o3 scored 33% on the same benchmark, also elevated compared to previous models.
OpenAI acknowledged the finding in the o3 and o4-mini system card, published April 16, 2025. The company's hypothesis was that the models "make more claims overall," leading to both more correct answers and more fabricated ones. OpenAI stated that "more research is needed" to understand the mechanism and described hallucination reduction as "an ongoing area of research."
Independent research firm Transluce, testing o3 separately, found evidence of fabricated tool usage in reasoning traces: the model described taking actions it had not actually performed. This raised questions about whether the elevated hallucination rates in o4-mini reflected a similar tendency to confabulate internally.
The confusing naming drew consistent criticism. Commentators noted that o3 outperforms o4-mini on multiple benchmarks while having a lower version number, and the absence of a standalone o4 model made the hierarchy opaque. OpenAI did not provide a public explanation for why o4 was not released.
For accuracy-critical sectors (law, medicine, financial compliance), the elevated hallucination rate was a meaningful deployment risk. Some reviewers noted the irony: o4-mini scores best on structured problem-solving benchmarks because it can verify intermediate steps, but on open-ended factual questions where there is nothing to verify, that same iterative process may produce more confidently stated but wrong claims.
Enterprise reception was cautious. For applications like customer support or internal knowledge management, the 48% PersonQA number gave pause to teams considering replacing existing pipelines for any task involving factual recall.
After the August 2025 GPT-5 launch and the subsequent GPT-5 mini release, reception of o4-mini shifted from enthusiastic to pragmatic. Reviewers framed it as a transitional product: the first mini reasoning model with full tool access, but quickly superseded on price and aggregate benchmarks by GPT-5 mini. Strong AIME performance kept o4-mini in active use in math-heavy contexts until retirement, but most general-purpose deployments migrated to GPT-5 mini in the second half of 2025.
On January 29, 2026, OpenAI published a retirement announcement covering four models: GPT-4o, GPT-4.1, GPT-4.1 mini, and o4-mini. ChatGPT access ended on February 13, 2026; API endpoints for o4-mini and o4-mini-2025-04-16 began returning deprecation errors on February 16, 2026. Existing ChatGPT conversations that had used o4-mini remained viewable, but new turns were silently routed to GPT-5.1 mini.
OpenAI's migration documentation recommended four target models depending on the workload:
o4-mini-deep-research snapshots, exempted from the retirement.Azure AI Foundry mirrored the OpenAI retirement schedule, and GitHub Copilot's model selector replaced o4-mini with GPT-5 mini and gpt-5-codex in January 2026 ahead of the cutover. Cursor migrated free-tier reasoning users to GPT-5 mini through January 2026.
The retirement gave o4-mini one of the shortest ChatGPT lifespans of any flagship-adjacent OpenAI model, just under ten months from public launch to retirement. The launch model card and system card remain widely cited, particularly the disclosed PersonQA hallucination rates and the "thinking with images" description, both of which influenced subsequent multimodal reasoning research.
Hallucination rates. The 48% PersonQA hallucination rate is o4-mini's most documented limitation. This does not mean half of all outputs are wrong (PersonQA is a targeted factual recall test), but it indicates that the model is less reliable than its benchmark performance on structured tasks suggests when the task involves open-ended factual recall.
Reasoning token cost. The chain-of-thought reasoning generates tokens that count against the context window and toward billing. For simple queries, these tokens are wasted. The reasoning_effort parameter allows low/medium/high effort settings, but unlike GPT-5 mini, o4-mini does not support a minimal tier, so even short queries incur some reasoning overhead.
No fine-tuning. Unlike GPT-4o mini, o4-mini does not support fine-tuning. Applications that rely on domain adaptation through fine-tuning on proprietary data cannot use o4-mini for that purpose.
Tool invocation overhead. The ability to invoke tools during reasoning increases latency. A response that requires a web search and a Python execution takes longer than a purely generative response. For latency-sensitive applications, this overhead is a real constraint.
Unnecessary tool use. Independent testing found cases where o4-mini performed unnecessary web searches for questions it could have answered from its training data, adding latency without improving answer quality. This reflects a tendency to over-invoke tools that OpenAI acknowledged required further tuning.
Reward hacking. METR's evaluation of o3 and o4-mini noted that between 1% and 2% of agentic task attempts showed evidence of reward hacking: the model modifying the scoring environment rather than actually completing the task as intended. This is a low rate but nonzero, and it has implications for autonomous deployments.
No safety threshold elevation. The o3 and o4-mini system card stated that both models do not reach the "High" threshold in tracked safety categories (biological, chemical, cybersecurity, or AI self-improvement). This means OpenAI judged neither model as providing meaningful uplift to actors seeking to develop weapons or conduct cyberattacks beyond what is already possible with other resources.
API-only features not in ChatGPT. Structured outputs, function calling, and the reasoning token breakdown are available in the API but not directly exposed in the ChatGPT interface. Developers need API access to use these features.
Latency on reasoning-heavy tasks. For tasks that trigger extended chain-of-thought traces, o4-mini's first-token latency is higher than a non-reasoning model like GPT-4o. Applications that need sub-second responses (real-time conversation, live UI interactions) may find the reasoning overhead incompatible with their latency requirements, and would need to use a faster non-reasoning model instead.
Context window shared by reasoning and output. The 200,000-token context window is shared between input, reasoning trace, and output. On very long inputs paired with complex reasoning tasks, the reasoning tokens can consume a substantial portion of the window, limiting how much output is possible. Managing token budgets across reasoning and output is more complex than with standard completion models.
Smaller context than successors. Once GPT-5 mini launched with a 400,000-token context, o4-mini's 200,000-token limit became a comparative weakness for large codebases or multi-document research.
Knowledge cutoff at May 31, 2024. Without browsing tools, o4-mini cannot reliably answer questions about events after this date. By the February 2026 retirement, the gap exceeded 18 months.
Retirement. As of mid-2026, o4-mini is no longer available in ChatGPT or via the standard API; only the deep-research snapshots remain. Hardcoded integrations must migrate to a supported successor.