<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0">
    <channel>
        <title>NerdLevelTech — AI Model Index</title>
        <link>https://nerdleveltech.com/models</link>
        <description>Published API prices, context windows, licences and editorial picks for the AI models we track. Facts are vendor-sourced with dated links; picks are labelled editorial; no benchmark score is published.</description>
        <lastBuildDate>Sun, 25 Oct 2026 12:38:04 GMT</lastBuildDate>
        <docs>https://validator.w3.org/feed/docs/rss2.html</docs>
        <generator>https://github.com/jpmonette/feed</generator>
        <language>en</language>
        <image>
            <title>NerdLevelTech — AI Model Index</title>
            <url>https://nerdleveltech.com/favicon.png</url>
            <link>https://nerdleveltech.com/models</link>
        </image>
        <copyright>All rights reserved 2026, Nerd Level Tech</copyright>
        <item>
            <title><![CDATA[Our picks: Claude Opus 5.5 (Best overall), GPT-6.1 Sol (Best value), DeepSeek V4.1 Flash (Best budget workhorse), GLM-5.3 Flash (Best open weights)]]></title>
            <link>https://nerdleveltech.com/models/leaderboard</link>
            <guid>https://nerdleveltech.com/models/leaderboard</guid>
            <pubDate>Sat, 24 Oct 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Best overall: Claude Opus 5.5 (Paid API) — Anthropic recommends it as the starting point for most workloads, and it backs that up with a 1M context, documented tool use and a $4/$20 price. The default we would ship first.

Best value: GPT-6.1 Sol (Paid API) — OpenAI positions it as near-Astra performance for complex coding at $2/$10, with the same 1.05M context. The best quality-per-dollar on this list.

Best budget workhorse: DeepSeek V4.1 Flash (Open weights) — The cheap agent pick: vision, a 1M-token context window and tool use at about $0.027 per task, half that off-peak. GLM-5.3 Flash is cheaper for text-only work; this is the one we would hand an agent.

Best open weights: GLM-5.3 Flash (Self-host or API) — The open-weight pick: a dedicated Hugging Face repo, an Ollama library, a 1M context and native image input at $0.15/$0.50 per 1M. The full GLM-5.3 costs about nine times more for the same window.]]></description>
            <category>picks</category>
            <category>editorial</category>
        </item>
        <item>
            <title><![CDATA[Kimi K3.1 Pro]]></title>
            <link>https://nerdleveltech.com/models/kimi-k3-1-pro</link>
            <guid>https://nerdleveltech.com/models/kimi-k3-1-pro</guid>
            <pubDate>Fri, 23 Oct 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[The 3T-class challenger that undercuts Opus

Moonshot’s flagship: a 3T-class open-weight model with a 1M context and native visual understanding, positioned for long-horizon coding and knowledge work. The weights are promised by November 4, 2026; the API lists at $2/$10 per 1M tokens.

Price: $2 input / $10 output per 1M tokens. About $0.20 for 50K input + 10K output tokens at the headline rates.
Context window: 1,000,000 tokens.
Licence: open weights announced, but not downloadable yet.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/kimi-k3-1-pro]]></description>
            <category>moonshot</category>
            <category>Moonshot AI</category>
            <category>open weights</category>
            <category>coding</category>
            <category>reasoning</category>
            <category>vision</category>
            <category>agentic</category>
        </item>
        <item>
            <title><![CDATA[MiniMax M3.1 Flash]]></title>
            <link>https://nerdleveltech.com/models/minimax-m3-1-flash</link>
            <guid>https://nerdleveltech.com/models/minimax-m3-1-flash</guid>
            <pubDate>Mon, 19 Oct 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[1M context and open weights at a Chinese price

MiniMax’s agentic Flash model: a 1M context, native multimodal input, adjustable thinking depth, open weights on Hugging Face, and $0.15/$0.60 per 1M tokens on the global API.

Price: $0.15 input / $0.6 output per 1M tokens. About $0.013 for 50K input + 10K output tokens at the headline rates.
Context window: 1,000,000 tokens.
Licence: open weights (downloadable).
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/minimax-m3-1-flash]]></description>
            <category>minimax</category>
            <category>MiniMax</category>
            <category>open weights</category>
            <category>agentic</category>
            <category>coding</category>
            <category>vision</category>
            <category>local</category>
        </item>
        <item>
            <title><![CDATA[DeepSeek V4.1 Flash — Best budget workhorse]]></title>
            <link>https://nerdleveltech.com/models/deepseek-v4-1-flash</link>
            <guid>https://nerdleveltech.com/models/deepseek-v4-1-flash</guid>
            <pubDate>Sat, 17 Oct 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Open weights at a fraction of frontier prices

DeepSeek’s Flash tier: a 1M-token context window, thinking on by default, vision input and open weights. Peak API pricing is $0.30/$1.20 per 1M tokens; off-peak is half.

Price: $0.3 input / $1.2 output per 1M tokens. About $0.027 for 50K input + 10K output tokens at the headline rates.
Context window: 1,000,000 tokens.
Licence: open weights (downloadable).
Pick: Best budget workhorse (Open weights) — The cheap agent pick: vision, a 1M-token context window and tool use at about $0.027 per task, half that off-peak. GLM-5.3 Flash is cheaper for text-only work; this is the one we would hand an agent.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/deepseek-v4-1-flash]]></description>
            <category>deepseek</category>
            <category>DeepSeek</category>
            <category>open weights</category>
            <category>agentic</category>
            <category>coding</category>
            <category>vision</category>
        </item>
        <item>
            <title><![CDATA[Nemotron 3.5 Super VL]]></title>
            <link>https://nerdleveltech.com/models/nemotron-3-5-super-vl</link>
            <guid>https://nerdleveltech.com/models/nemotron-3-5-super-vl</guid>
            <pubDate>Thu, 15 Oct 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[NVIDIA’s open vision-language model with a 1M window

NVIDIA’s open vision-language model: a 1,048,576-token context, text and image input, and $0.15/$0.50 per 1M tokens on the hosted API. The public Hugging Face release is the NVFP4 build.

Price: $0.15 input / $0.5 output per 1M tokens. About $0.013 for 50K input + 10K output tokens at the headline rates.
Context window: 1,048,576 tokens.
Licence: open weights (downloadable).
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/nemotron-3-5-super-vl]]></description>
            <category>nvidia</category>
            <category>NVIDIA</category>
            <category>open weights</category>
            <category>vision</category>
            <category>agentic</category>
            <category>local</category>
            <category>tool-use</category>
        </item>
        <item>
            <title><![CDATA[Gemini 4 Argon]]></title>
            <link>https://nerdleveltech.com/models/gemini-4-argon</link>
            <guid>https://nerdleveltech.com/models/gemini-4-argon</guid>
            <pubDate>Tue, 13 Oct 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Google’s flagship: 1M context and long decodes

Google’s flagship reasoning model, built for software engineering, long-horizon agents and long decodes. One 1,048,576-token window covers input and output; a standard call outputs up to 65,536 tokens, with up to 1M in long-decode mode.

Price: $2 input / $10 output per 1M tokens. About $0.20 for 50K input + 10K output tokens at the headline rates.
Context window: 1,048,576 tokens.
Licence: proprietary.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/gemini-4-argon]]></description>
            <category>google</category>
            <category>Google DeepMind</category>
            <category>proprietary</category>
            <category>agentic</category>
            <category>coding</category>
            <category>reasoning</category>
            <category>vision</category>
        </item>
        <item>
            <title><![CDATA[Claude Haiku 5.5]]></title>
            <link>https://nerdleveltech.com/models/claude-haiku-5-5</link>
            <guid>https://nerdleveltech.com/models/claude-haiku-5-5</guid>
            <pubDate>Wed, 07 Oct 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Cheap, fast, and good enough for the boring 80%

Anthropic’s cheapest current model, built for high-volume, latency-sensitive work such as classification, extraction and routing. $0.10/$0.50 per 1M tokens up to 100K-token prompts.

Price: $0.1 input / $0.5 output per 1M tokens. About $0.010 for 50K input + 10K output tokens at the headline rates.
Context window: 1,000,000 tokens.
Licence: proprietary.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/claude-haiku-5-5]]></description>
            <category>anthropic</category>
            <category>Anthropic</category>
            <category>proprietary</category>
            <category>high-volume</category>
            <category>vision</category>
            <category>tool-use</category>
        </item>
        <item>
            <title><![CDATA[Mistral Large 4]]></title>
            <link>https://nerdleveltech.com/models/mistral-large-4</link>
            <guid>https://nerdleveltech.com/models/mistral-large-4</guid>
            <pubDate>Tue, 06 Oct 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Europe’s 1T challenger

Mistral’s 1.05T-parameter mixture-of-experts model with 52B active parameters, a 1M context, text and image input, and an open-weight release promised after the API preview. API pricing is $1.36/$4.18 per 1M tokens.

Price: $1.36 input / $4.18 output per 1M tokens. About $0.11 for 50K input + 10K output tokens at the headline rates.
Context window: 1,000,000 tokens.
Licence: open weights announced, but not downloadable yet.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/mistral-large-4]]></description>
            <category>mistral</category>
            <category>Mistral AI</category>
            <category>open weights</category>
            <category>agentic</category>
            <category>coding</category>
            <category>reasoning</category>
            <category>vision</category>
        </item>
        <item>
            <title><![CDATA[GPT-6.1 Sol — Best value]]></title>
            <link>https://nerdleveltech.com/models/gpt-6-1-sol</link>
            <guid>https://nerdleveltech.com/models/gpt-6-1-sol</guid>
            <pubDate>Tue, 29 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Near-Astra quality at a fifth of the price

OpenAI positions GPT-6.1 Sol as near-Astra performance for complex coding and computer use, at $2/$10 per 1M tokens with the same 1,050,000-token window (922K max input + 128K max output).

Price: $2 input / $10 output per 1M tokens. About $0.20 for 50K input + 10K output tokens at the headline rates.
Context window: 1,050,000 tokens.
Licence: proprietary.
Pick: Best value (Paid API) — OpenAI positions it as near-Astra performance for complex coding at $2/$10, with the same 1.05M context. The best quality-per-dollar on this list.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/gpt-6-1-sol]]></description>
            <category>openai</category>
            <category>OpenAI</category>
            <category>proprietary</category>
            <category>coding</category>
            <category>agentic</category>
            <category>vision</category>
            <category>computer-use</category>
        </item>
        <item>
            <title><![CDATA[Claude Sonnet 5.5]]></title>
            <link>https://nerdleveltech.com/models/claude-sonnet-5-5</link>
            <guid>https://nerdleveltech.com/models/claude-sonnet-5-5</guid>
            <pubDate>Mon, 28 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[The speed/intelligence balance most teams should start with

Anthropic’s middle tier: the same 1M-token context window as Opus, with output capped at 128K, at half the price, with a high default effort and the fastest latency of the current line-up.

Price: $2 input / $10 output per 1M tokens. About $0.20 for 50K input + 10K output tokens at the headline rates.
Context window: 1,000,000 tokens.
Licence: proprietary.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/claude-sonnet-5-5]]></description>
            <category>anthropic</category>
            <category>Anthropic</category>
            <category>proprietary</category>
            <category>agentic</category>
            <category>coding</category>
            <category>vision</category>
            <category>tool-use</category>
        </item>
        <item>
            <title><![CDATA[Claude Opus 5.5 — Best overall]]></title>
            <link>https://nerdleveltech.com/models/claude-opus-5-5</link>
            <guid>https://nerdleveltech.com/models/claude-opus-5-5</guid>
            <pubDate>Tue, 22 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Anthropic's default for long-running agentic coding

Anthropic’s current flagship, positioned for long-running agentic coding and knowledge work. Adaptive thinking is always on, with a medium default effort that you can raise per task.

Price: $4 input / $20 output per 1M tokens. About $0.40 for 50K input + 10K output tokens at the headline rates.
Context window: 1,000,000 tokens.
Licence: proprietary.
Pick: Best overall (Paid API) — Anthropic recommends it as the starting point for most workloads, and it backs that up with a 1M context, documented tool use and a $4/$20 price. The default we would ship first.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/claude-opus-5-5]]></description>
            <category>anthropic</category>
            <category>Anthropic</category>
            <category>proprietary</category>
            <category>agentic</category>
            <category>coding</category>
            <category>reasoning</category>
            <category>tool-use</category>
            <category>vision</category>
        </item>
        <item>
            <title><![CDATA[GPT-6 Luna]]></title>
            <link>https://nerdleveltech.com/models/gpt-6-luna</link>
            <guid>https://nerdleveltech.com/models/gpt-6-luna</guid>
            <pubDate>Tue, 22 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[The cheapest model OpenAI ships

OpenAI’s high-volume tier at a base rate of $0.10/$0.50 per 1M tokens, with the same 1,050,000-token window (922K max input + 128K max output) as the rest of the GPT-6 family.

Price: $0.1 input / $0.5 output per 1M tokens. About $0.010 for 50K input + 10K output tokens at the headline rates.
Context window: 1,050,000 tokens.
Licence: proprietary.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/gpt-6-luna]]></description>
            <category>openai</category>
            <category>OpenAI</category>
            <category>proprietary</category>
            <category>high-volume</category>
            <category>vision</category>
            <category>tool-use</category>
        </item>
        <item>
            <title><![CDATA[Grok 4.7]]></title>
            <link>https://nerdleveltech.com/models/grok-4-7</link>
            <guid>https://nerdleveltech.com/models/grok-4-7</guid>
            <pubDate>Mon, 21 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[xAI’s frontier model with a 500K context

xAI’s frontier model for coding, agentic tasks and knowledge work: a 500,000-token context, text and image input, no stated output limit, and reasoning from low to xhigh.

Price: $2 input / $6 output per 1M tokens. About $0.16 for 50K input + 10K output tokens at the headline rates.
Context window: 500,000 tokens.
Licence: proprietary.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/grok-4-7]]></description>
            <category>xai</category>
            <category>xAI</category>
            <category>proprietary</category>
            <category>agentic</category>
            <category>coding</category>
            <category>reasoning</category>
            <category>vision</category>
        </item>
        <item>
            <title><![CDATA[GPT-6 Astra]]></title>
            <link>https://nerdleveltech.com/models/gpt-6-astra</link>
            <guid>https://nerdleveltech.com/models/gpt-6-astra</guid>
            <pubDate>Thu, 03 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[OpenAI’s most capable model for demanding work

OpenAI’s flagship: a 1,050,000-token window (922K max input + 128K max output), text and image input, and reasoning effort from low to max. Prompts above 272K input tokens are billed at 2× input.

Price: $10 input / $50 output per 1M tokens. About $1.00 for 50K input + 10K output tokens at the headline rates.
Context window: 1,050,000 tokens.
Licence: proprietary.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/gpt-6-astra]]></description>
            <category>openai</category>
            <category>OpenAI</category>
            <category>proprietary</category>
            <category>agentic</category>
            <category>coding</category>
            <category>reasoning</category>
            <category>vision</category>
            <category>computer-use</category>
        </item>
        <item>
            <title><![CDATA[Gemini 3.8 Flash]]></title>
            <link>https://nerdleveltech.com/models/gemini-3-8-flash</link>
            <guid>https://nerdleveltech.com/models/gemini-3-8-flash</guid>
            <pubDate>Wed, 02 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Google’s fast model for agents and long code

Google’s Flash tier for long-horizon software engineering and agent workflows: 1,048,576-token input, 65,536-token output, text, image, video, audio and PDF input, and a free tier.

Price: $0.75 input / $3.75 output per 1M tokens. About $0.075 for 50K input + 10K output tokens at the headline rates.
Context window: 1,048,576 tokens.
Licence: proprietary.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/gemini-3-8-flash]]></description>
            <category>google</category>
            <category>Google DeepMind</category>
            <category>proprietary</category>
            <category>agentic</category>
            <category>coding</category>
            <category>vision</category>
            <category>tool-use</category>
        </item>
        <item>
            <title><![CDATA[Muse Spark 1.3]]></title>
            <link>https://nerdleveltech.com/models/muse-spark-1-3</link>
            <guid>https://nerdleveltech.com/models/muse-spark-1-3</guid>
            <pubDate>Wed, 02 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Meta’s agentic coder with a 1M window

Meta’s agentic and coding model: a 1,048,576-token context, text, image, video, audio and PDF input, and standard pricing of $1.25/$4.25 per 1M tokens. Meta also publishes an open-weight release, Muse Spark Open.

Price: $1.25 input / $4.25 output per 1M tokens. About $0.10 for 50K input + 10K output tokens at the headline rates.
Context window: 1,048,576 tokens.
Licence: open weights (downloadable).
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/muse-spark-1-3]]></description>
            <category>meta</category>
            <category>Meta</category>
            <category>open weights</category>
            <category>agentic</category>
            <category>coding</category>
            <category>vision</category>
            <category>local</category>
        </item>
        <item>
            <title><![CDATA[Claude Fable 5.1]]></title>
            <link>https://nerdleveltech.com/models/claude-fable-5-1</link>
            <guid>https://nerdleveltech.com/models/claude-fable-5-1</guid>
            <pubDate>Tue, 01 Sep 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Premium reasoning when Opus is not enough

Anthropic’s most demanding-work model: adaptive thinking always on, a 1M-token context window with output capped at 128K, positioned for long-horizon agentic work where Opus at higher effort still falls short.

Price: $10 input / $50 output per 1M tokens. About $1.00 for 50K input + 10K output tokens at the headline rates.
Context window: 1,000,000 tokens.
Licence: proprietary.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/claude-fable-5-1]]></description>
            <category>anthropic</category>
            <category>Anthropic</category>
            <category>proprietary</category>
            <category>reasoning</category>
            <category>agentic</category>
            <category>coding</category>
            <category>vision</category>
        </item>
        <item>
            <title><![CDATA[GLM-5.3 Flash — Best open weights]]></title>
            <link>https://nerdleveltech.com/models/glm-5-3-flash</link>
            <guid>https://nerdleveltech.com/models/glm-5-3-flash</guid>
            <pubDate>Wed, 26 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[The cheap multimodal GLM

Z.ai’s cheap multimodal coding model: 320B total parameters with 18B active, a 1M context, native image input, and $0.15/$0.50 per 1M tokens. Same post-training family as GLM-5.3, built to be the everyday tier.

Price: $0.15 input / $0.5 output per 1M tokens. About $0.013 for 50K input + 10K output tokens at the headline rates.
Context window: 1,000,000 tokens.
Licence: open weights (downloadable).
Pick: Best open weights (Self-host or API) — The open-weight pick: a dedicated Hugging Face repo, an Ollama library, a 1M context and native image input at $0.15/$0.50 per 1M. The full GLM-5.3 costs about nine times more for the same window.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/glm-5-3-flash]]></description>
            <category>zai</category>
            <category>Z.ai</category>
            <category>open weights</category>
            <category>coding</category>
            <category>vision</category>
            <category>local</category>
            <category>high-volume</category>
        </item>
        <item>
            <title><![CDATA[GLM-5.3]]></title>
            <link>https://nerdleveltech.com/models/glm-5-3</link>
            <guid>https://nerdleveltech.com/models/glm-5-3</guid>
            <pubDate>Tue, 18 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[The open-weight coder that keeps closing the gap

Z.ai’s flagship coding model: a 1M-token context window, text-only input, reasoning always on with low/high/max effort. Same base as GLM-5.2, with the gains coming from post-training.

Price: $1.4 input / $4.4 output per 1M tokens. About $0.11 for 50K input + 10K output tokens at the headline rates.
Context window: 1,000,000 tokens.
Licence: open weights (downloadable).
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/glm-5-3]]></description>
            <category>zai</category>
            <category>Z.ai</category>
            <category>open weights</category>
            <category>coding</category>
            <category>reasoning</category>
            <category>local</category>
        </item>
        <item>
            <title><![CDATA[Qwen 3.8 27B]]></title>
            <link>https://nerdleveltech.com/models/qwen-3-8-27b</link>
            <guid>https://nerdleveltech.com/models/qwen-3-8-27b</guid>
            <pubDate>Fri, 14 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[The local 27B that fits on a laptop

Alibaba’s open 27B model: a 1M context on the API, text and image input, $0.20/$2.50 per 1M tokens, and an 18GB Ollama build that runs a 256K window.

Price: $0.2 input / $2.5 output per 1M tokens. About $0.035 for 50K input + 10K output tokens at the headline rates.
Context window: 1,000,000 tokens.
Licence: open weights (downloadable).
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/qwen-3-8-27b]]></description>
            <category>alibaba</category>
            <category>Alibaba (Qwen)</category>
            <category>open weights</category>
            <category>local</category>
            <category>vision</category>
            <category>coding</category>
            <category>high-volume</category>
        </item>
        <item>
            <title><![CDATA[DeepSeek V4 Pro]]></title>
            <link>https://nerdleveltech.com/models/deepseek-v4-pro</link>
            <guid>https://nerdleveltech.com/models/deepseek-v4-pro</guid>
            <pubDate>Thu, 13 Aug 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[The Pro tier for hard reasoning, not everyday coding

DeepSeek’s Pro tier: a 1M-token context window, thinking on by default and open weights. No vision. Peak pricing is $0.66/$1.98 per 1M tokens; off-peak is half.

Price: $0.66 input / $1.98 output per 1M tokens. About $0.053 for 50K input + 10K output tokens at the headline rates.
Context window: 1,000,000 tokens.
Licence: open weights (downloadable).
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/deepseek-v4-pro]]></description>
            <category>deepseek</category>
            <category>DeepSeek</category>
            <category>open weights</category>
            <category>reasoning</category>
            <category>coding</category>
            <category>local</category>
        </item>
        <item>
            <title><![CDATA[Composer 2.5]]></title>
            <link>https://nerdleveltech.com/models/composer-2-5</link>
            <guid>https://nerdleveltech.com/models/composer-2-5</guid>
            <pubDate>Mon, 18 May 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Cursor’s own coding model, built into the editor

Cursor’s in-house agent model: a 200K context, agentic tool use, a fast interactive tier and a lower cost per token than the frontier APIs. It is part of Cursor’s plans rather than a per-token public API.

Price: Cursor plan.
Context window: 200,000 tokens.
Licence: proprietary.
Verified: 2026-10-24. Source: https://nerdleveltech.com/models/composer-2-5]]></description>
            <category>cursor</category>
            <category>Cursor</category>
            <category>proprietary</category>
            <category>agentic</category>
            <category>coding</category>
        </item>
    </channel>
</rss>