Revision History

Correction: re-checked the third-party host prices on October 30, 2026. Together AI now charges DeepSeek's peak-hour rate for V4-Pro at all hours (was $1.74 / $3.48 on August 1) and lists DeepSeek V4.1-Flash; Groq withdrew Llama 3.1 8B, Llama 3.3 70B and Qwen 3.6 27B from non-enterprise users and lists Qwen 3.8 27B in preview; Amazon Bedrock prices are unchanged and now include GPT-6 Astra.

Correction: re-checked every first-party provider on October 30, 2026; added GPT-6.1 Terra, OpenAI 12-hour caching, Gemini 4 Argon, Mistral Large 4's new $0.70 / $2.10 price, Qwen3.8-Max, GLM-5.3, Kimi K3.1 Pro, MiniMax M3 and Muse Wonder; removed Qwen3-Max and the deprecated Gemini 3.7 Flash and 3.5 Flash rows; noted the xAI rename to SpaceXSI and its 20% batch discount.

October 13, 2026 price re-check: added GPT-6 Sol, GPT-6.1 Sol, GPT-6 Luna, Claude Opus 5.5, Sonnet 5.5, Haiku 5.5, Grok 4.7, Mistral Large 4, and DeepSeek V4.1-Flash; removed retired DeepSeek V4-Flash and Mistral Magistral/Devstral 2 rows; marked deprecations; Muse Spark now sourced to Meta's pricing page.

Correction: the Ultrafast tier was described in the future tense as an August 13 GPT-5.6 Sol preview with no published price; it shipped for GPT-6 Astra on September 29, 2026 at $60 / $6 / $75 / $300 per 1M tokens (6x Standard). Added the September 29 timeline row and four sources.

Link Claude Fable 5.1 and Claude Mythos 5.1 to their new pages

Correction: GPT-6 Astra API pricing added ($10/$1/$12.50/$50, Fast mode 2x, Bedrock); GPT-5.6 Sol promotional cut of August 21 ($4/$0.40/$20 through at least Nov 21) replaces the stale $5/$30; DeepSeek peak/off-peak repricing in force since August 16; Claude Sonnet 5 September 1 increase cancelled; Claude Fable 5.1, Gemini 3.6/3.7/3.8 Flash promotional rate, and Grok 4.6 added; every OpenAI row re-verified September 3, 2026.

Refresh DeepSeek peak-pricing as-of date to Aug 5 (re-verified on the pricing page) and add Bloomberg's Aug 4 cost-per-task comparison

Correction: Qwen-Plus bills chain-of-thought output separately at $4.00 and $12.00 per 1M by context band, so the listed output rate understated the cost of a reasoning-mode request

Correction: Amazon Nova Micro and Nova Pro rates did not match AWS's published pricing, a superlative labelled an input rate as blended, and the claim that DeepSeek V4-Pro cost more to evaluate because it emitted more tokens is contradicted by Artificial Analysis, which records fewer; also corrects the fast-mode multiplier for gpt-5.5, two cache-discount figures, and the description of Simon Willison's test prompt

Full re-verification of every price against provider pricing pages on August 1, 2026, with an explicit prices-as-of date. Adds the July 30 GPT-5.6 Luna (-80%) and Terra (-20%) cuts and the Fast mode rename, plus Claude Opus 5 / Sonnet 5 / Fable 5, Gemini 3.6 Flash, Grok 4.5, Kimi K3, GLM, and hosted open-weights rates; corrects Qwen3-Max and Grok 4.3; adds sections on cost per task vs per-token rates, cache economics, long-context surcharges, service tiers, and why DeepSeek peak pricing is not in force.

Added 6 contextual internal links

No summary provided