# Coding Models Index Canonical page: https://amplifying.ai/models Latest release details checked October 2, 2026. Benchmark scores and usage observations: August 26, 2026, with older figures marked individually. These are curated snapshots, not live prices or scores. Prices are USD per million tokens (input / output), with the provider and tier noted. Prices without a checked date retain the August snapshot; do not treat them as current quotes. Context tiers, caching, and other terms can affect the bill. Unscored means no comparable result in this snapshot, not a score of zero. This Markdown companion lists releases, prices, coding-index observations, and sources. See the HTML page for the composite scoring method and additional adoption comparisons. | Model | Provider | Released | Input / Output USD | Pricing date | August coding index | | --- | --- | --- | --- | --- | --- | | GPT-6.1 Sol | OpenAI | 2026-09-29 | 2 / 10 | Checked October 2, 2026 | Unscored | | Claude Sonnet 5.5 | Anthropic | 2026-09-28 | 2 / 10 | Checked October 2, 2026 | Unscored | | Claude Opus 5.5 | Anthropic | 2026-09-22 | 4 / 20 | Checked October 2, 2026 | Unscored | | GPT-6 Sol | OpenAI | 2026-09-22 | 2 / 10 | Checked October 2, 2026 | Unscored | | GPT-6 Luna | OpenAI | 2026-09-22 | 0.1 / 0.5 | Checked October 2, 2026 | Unscored | | Grok 4.7 | xAI | 2026-09-21 | 2 / 6 | Checked October 2, 2026 | Unscored | | DeepSeek V4.1 Flash | DeepSeek | 2026-09-10 | 0.3 / 1.2 | Checked October 2, 2026 | Unscored | | GPT-6 Astra | OpenAI | 2026-09-03 | 10 / 50 | Checked October 2, 2026 | Unscored | | Gemini 3.8 Flash | Google | 2026-09-02 | 0.75 / 3.75 | Checked October 2, 2026 | Unscored | | Claude Fable 5.1 | Anthropic | 2026-09-01 | 10 / 50 | Checked October 2, 2026 | Unscored | | GLM 5.3 Flash (Ox Alpha) | Z.ai (revealed Aug 26) | 2026-08-20 | 0.075 / 0.25 | Snapshot August 26, 2026 | Unscored | | GLM 5.3 | Z.ai | 2026-08-14 | 1.4 / 4.4 | Snapshot August 26, 2026 | 74.8 (secondhand source) | | Gemini 3.7 Flash | Google | 2026-08-13 | 0.375 / 1.875 | Snapshot August 26, 2026 | 76.1 | | Grok 4.6 | xAI | 2026-08-12 | 2 / 6 | Snapshot August 26, 2026 | 76.8 | | DeepSeek V4 Pro | DeepSeek | 2026-08-12 | 1.32 / 3.96 | Checked October 2, 2026 | 68.8 (secondhand source) | | Muse Spark 1.2 | Meta | 2026-08-05 | 1.25 / 4.25 | Snapshot August 26, 2026 | 72.2 (secondhand source) | | Qwen 3.8 Max | Alibaba | 2026-08-03 | 2 / 6 | Snapshot August 26, 2026 | 71.8 (secondhand source) | | DeepSeek V4 Flash | DeepSeek | 2026-07-31 | 0.05 / 0.1 | Snapshot August 26, 2026 | 69.1 (secondhand source) | | Claude Opus 5 | Anthropic | 2026-07-24 | 5 / 25 | Snapshot August 26, 2026 | 78 | | Kimi K3 | Moonshot AI | 2026-07-16 | 3 / 15 | Snapshot August 26, 2026 | 76.2 | | GPT-5.6 Sol | OpenAI | 2026-07-09 | 4 / 20 | Checked October 2, 2026 | 78.3 | | GPT-5.6 Terra | OpenAI | 2026-07-09 | 2 / 12 | Snapshot August 26, 2026 | 76.7 | | GPT-5.6 Luna | OpenAI | 2026-07-09 | 0.2 / 1.2 | Snapshot August 26, 2026 | 71.5 (secondhand source) | | Claude Sonnet 5 | Anthropic | 2026-06-30 | 2 / 10 | Snapshot August 26, 2026 | 71.5 | | Claude Fable 5 | Anthropic | 2026-06-09 | 10 / 50 | Snapshot August 26, 2026 | 76.5 | | MiniMax M3 | MiniMax | 2026-05-31 | 0.3 / 1.2 | Checked October 2, 2026 | 58.6 (secondhand source) | | Composer 2.5 | Cursor | 2026-05-18 | 0.5 / 2.5 | Checked October 2, 2026 | Unscored | | GPT-5.5 | OpenAI | 2026-04-24 | 5 / 30 | Snapshot August 26, 2026 | 74.9 | | Gemini 3.1 Pro | Google | 2026-02-19 | 2 / 12 | Snapshot August 26, 2026 | 68.8 (secondhand source) | | Claude Haiku 4.5 | Anthropic | 2025-10-15 | 1 / 5 | Snapshot August 26, 2026 | Unscored | ## GPT-6.1 Sol The updated Sol model for coding and computer use, with a 1.05M-token context window. Use the Responses API for tool calling. Price date: October 2, 2026. Pricing note: standard up to 272K input; cache reads $0.10; above 272K: $4 input / $15 output Released after the August 26 benchmark snapshot. No comparable score has been added here. - [Model and pricing](https://developers.openai.com/api/docs/models/gpt-6.1-sol) - [Release log](https://developers.openai.com/api/docs/changelog) ## Claude Sonnet 5.5 A faster, lower-cost alternative to Opus 5.5, with a 1M-token context window. Price date: October 2, 2026. Pricing note: standard API; cache reads $0.20 Released after the August 26 benchmark snapshot. No comparable score has been added here. - [Model, release date, and pricing](https://platform.claude.com/docs/en/models/sonnet-5-5/overview) ## Claude Opus 5.5 The newest Opus release. Standard input and output tokens cost 20% less than Opus 5. Price date: October 2, 2026. Pricing note: standard API; cache reads $0.20 Not scored in the August 26 benchmark snapshot. Release details are verified separately; no score or adoption figure has been inferred. - [Release and pricing](https://www.anthropic.com/claude-opus-5-5) ## GPT-6 Sol The original GPT-6 Sol release. GPT-6.1 Sol is the newer version at the same standard input and output rates, with a lower cache-read price. Price date: October 2, 2026. Pricing note: standard; input up to 272K Not scored in the August 26 benchmark snapshot. Release details are verified separately; no score or adoption figure has been inferred. - [Model and pricing](https://developers.openai.com/api/docs/models/gpt-6-sol) - [Release log](https://developers.openai.com/api/docs/changelog) ## GPT-6 Luna The lower-cost GPT-6 option for routine work and high-volume agent tasks. Price date: October 2, 2026. Pricing note: standard; input up to 272K Not scored in the August 26 benchmark snapshot. Release details are verified separately; no score or adoption figure has been inferred. - [Model and pricing](https://developers.openai.com/api/docs/models/gpt-6-luna) - [Release log](https://developers.openai.com/api/docs/changelog) ## Grok 4.7 A 500K-context model with configurable reasoning. Its faster variant is available through Cursor and Grok Build. Price date: October 2, 2026. Pricing note: standard; prompts below 200K Not scored in the August 26 benchmark snapshot. Release details are verified separately; no score or adoption figure has been inferred. - [Release and pricing](https://docs.x.ai/developers/release-notes) - [Model details](https://docs.x.ai/developers/grok-4-7) ## DeepSeek V4.1 Flash An open-weights model with native image input and 1M context. The deepseek-flash API alias serves V4.1; legacy V4 Flash IDs also route here. Price date: October 2, 2026. Pricing note: peak, uncached; off-peak $0.15 input / $0.60 output; cache hits $0.006 peak / $0.003 off-peak Released after the August 26 benchmark snapshot. Provider-reported results are not mixed into the historical score. - [Release log](https://api-docs.deepseek.com/updates/) - [API pricing and aliases](https://api-docs.deepseek.com/quick_start/pricing/) - [Weights and MIT license](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash) ## GPT-6 Astra OpenAI’s most capable broadly available model, combining coding, computer use, and research in long-running tasks. Price date: October 2, 2026. Pricing note: standard; input up to 272K Not scored in the August 26 benchmark snapshot. Release details are verified separately; no score or adoption figure has been inferred. - [Model and pricing](https://developers.openai.com/api/docs/models/gpt-6-astra) - [Release log](https://developers.openai.com/api/docs/changelog) ## Gemini 3.8 Flash Google’s latest Flash model for software engineering and agent workflows. The separate Cyber variant has restricted access. Price date: October 2, 2026. Pricing note: standard promo through Dec 31, 2026; $1.50 input / $7.50 output from Jan 1, 2027 Not scored in the August 26 benchmark snapshot. Release details are verified separately; no score or adoption figure has been inferred. - [Release](https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/) - [API pricing](https://ai.google.dev/gemini-api/docs/pricing) ## Claude Fable 5.1 Generally available for coding and knowledge work. The related Mythos 5.1 model is limited to trusted-access programs. Price date: October 2, 2026. Pricing note: standard API Not scored in the August 26 benchmark snapshot. Release details are verified separately; no score or adoption figure has been inferred. - [Release and pricing](https://www.anthropic.com/claude-fable-and-mythos-5-1) - [Release date](https://www.anthropic.com/claude/fable) ## GLM 5.3 Flash (Ox Alpha) Released as Ox Alpha before being identified as GLM 5.3 Flash. Volume and availability notes refer to the August snapshot. Description from the August snapshot. Price date: August 26, 2026. Pricing note: free during preview Coding index: not captured in August No AA or arena score was captured in the August 26 snapshot. ## GLM 5.3 Top-10 coding at a fraction of frontier pricing; 5.2 weights are open, 5.3 Flash weights promised. Description from the August snapshot. Price date: August 26, 2026. Coding index: max effort Coding index was obtained from a public mirror, not the primary board. ## Gemini 3.7 Flash August's fastest riser: near-frontier coding at Flash prices. Description from the August snapshot. Price date: August 26, 2026. Pricing note: 75% promo Coding index: high effort ## Grok 4.6 Cheapest model in the 76+ coding tier; AA flags it for cost efficiency. Description from the August snapshot. Price date: August 26, 2026. Coding index: high effort ## DeepSeek V4 Pro DeepSeek's reasoning flagship, refreshed mid-August. Description from the August snapshot. Price date: October 2, 2026. Availability: DeepSeek continues to serve V4 Pro after its previously announced September 14 retirement date. Pricing note: peak, uncached; off-peak $0.66 input / $1.98 output Coding index was obtained from a public mirror, not the primary board. - [Current API pricing and availability](https://api-docs.deepseek.com/quick_start/pricing/) ## Muse Spark 1.2 Meta's return to the coding table, ahead of Sonnet 5 on the Coding Index. Description from the August snapshot. Price date: August 26, 2026. Coding index was obtained from a public mirror, not the primary board. ## Qwen 3.8 Max Alibaba's frontier bid; the 2.4T-parameter open-weights sibling shipped Aug 12. Description from the August snapshot. Price date: August 26, 2026. Coding index: preview Coding index was obtained from a public mirror, not the primary board. ## DeepSeek V4 Flash The volume king of coding: usage spiked ~570% after the July 31 agent-tuned retrain. Description from the August snapshot. Price date: August 26, 2026. Availability: Retired on the DeepSeek API. Its legacy model ID now serves V4.1 Flash at V4.1 prices; the figures below are historical. Pricing note: approx; tiered Coding index was obtained from a public mirror, not the primary board. - [Retirement and alias routing](https://api-docs.deepseek.com/quick_start/pricing/) ## Claude Opus 5 #1 on AA's Intelligence Index; Claude Code + Opus 5 also tops their Coding Agent Index. Description from the August snapshot. Price date: August 26, 2026. Coding index: max effort ## Kimi K3 The open-weights frontier for coding, within ~2 points of Opus 5. Description from the August snapshot. Price date: August 26, 2026. Coding index: max effort ## GPT-5.6 Sol Highest Coding Index score in the August 26 snapshot. Description from the August snapshot. Price date: October 2, 2026. Pricing note: standard promo at least through Nov 21, 2026; input up to 272K Coding index: xhigh effort - [API pricing](https://developers.openai.com/api/docs/models/gpt-5.6-sol) ## GPT-5.6 Terra Coding-skewed 5.6 sibling: top-5 coding despite mid-pack general intelligence. Description from the August snapshot. Price date: August 26, 2026. Coding index: max effort ## GPT-5.6 Luna The volume play: near-Sonnet coding scores at commodity pricing. Description from the August snapshot. Price date: August 26, 2026. Coding index was obtained from a public mirror, not the primary board. ## Claude Sonnet 5 The workhorse mid-tier for agentic coding at a fifth of Fable pricing. Description from the August snapshot. Price date: August 26, 2026. ## Claude Fable 5 The hard-task tier: #2 on intelligence, priciest mainstream model per token. Description from the August snapshot. Price date: August 26, 2026. Coding index: max effort ## MiniMax M3 Open-weights long-horizon agent model with a 1M context at commodity cost. Description from the August snapshot. Price date: October 2, 2026. Pricing note: standard up to 512K input; above 512K: $0.60 input / $2.40 output; priority 1.5× Coding index was obtained from a public mirror, not the primary board. - [API pricing](https://platform.minimax.io/docs/pricing/overview) ## Composer 2.5 Cursor’s own coding model. Standard and Fast variants have different token prices. Description from the August snapshot. Price date: October 2, 2026. Pricing note: Cursor standard; default Fast variant is $3 input / $15 output Coding index: not captured in August No comparable independent score was captured in the August 26 snapshot. - [Cursor pricing and variants](https://cursor.com/docs/models/cursor-composer-2-5) ## GPT-5.5 The previous OpenAI flagship, still top-10 while the 5.6 family displaces it. Description from the August snapshot. Price date: August 26, 2026. Coding index: xhigh effort ## Gemini 3.1 Pro Google's Pro tier, now behind its own 3.7 Flash on both AA indexes. Description from the August snapshot. Price date: August 26, 2026. Coding index was obtained from a public mirror, not the primary board. ## Claude Haiku 4.5 Cheap, fast Claude tier for high-volume subagent and edit loops; aging against the 2026 field. Description from the August snapshot. Price date: August 26, 2026. Coding index: not captured in August No comparable score was captured in the August 26 snapshot. This is not a claim about today’s benchmark coverage. [Coding Agents Index: publicly attributed agent activity](https://amplifying.ai/coding-agents) [Decision Models Index: typed decisions, open implementations, and platforms](https://amplifying.ai/decision-models) [All agent-readable resources](https://amplifying.ai/llms.txt)