Qwen3.6vsComposer 2.5
Qwen3.6 | Composer 2.5 | |
|---|---|---|
| Specifications | ||
ParametersA rough measure of how big the model is. More parameters usually means more capable and more expensive to run, though it is a poor guide on its own — a smaller, newer model often beats a larger, older one. | 35B | — |
Context windowHow much text the model can hold in mind at once — your question, any documents you attach, the conversation so far, and its own reply. Go past it and the earliest part falls out of view. | 256k | — |
| API pricingUSD per 1M tokens · lower wins | ||
Input priceWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it. | $0.375 | — |
Output priceWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one. | $2.25 | — |
Cheapest inputLowest input rate across third-party providers, excluding the lab itself. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | $0.05Darkbloom | — |
Cheapest outputLowest output rate across third-party providers, excluding the lab itself. May come from a different provider than the cheapest input. | $0.70Darkbloom | — |
| Benchmarks | ||
SWE-Bench ProAgentic coding — Can the AI fix real bugs in real software? It's handed actual problems from open-source projects and has to write code that genuinely solves them. Higher is better. | 49.5% | 54% |
SWE-Bench MultilingualMultilingual coding — Like SWE-Bench, but the coding problems span many programming languages, not just one. Tests how broadly the AI can code. Higher is better. | 67.2% | 79.8% |
Terminal-Bench 2.0Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? (Version 2.0 of the test.) Higher is better. | 51.5% | 69.3% |
| BenchmarksPublished by one model only | ||
SWE-Bench VerifiedCoding — Real coding tasks pulled from open-source projects — the AI has to find and fix actual bugs. A human-checked version of the original SWE-Bench. Higher is better. | 73.4% | — |
DeepSWE 1.0Agentic coding — Artificial Analysis' independent test of deep, agentic software-engineering work — the AI has to plan and carry out substantial coding tasks end to end. Higher is better. | — | 18% |
Next.js EvalsNext.js coding — Vercel's open eval of how well AI coding agents build and migrate real Next.js apps — measured as the share of tasks the agent completes successfully. Higher is better. | — | 81% |
Terminal-Bench 2.1Agentic terminal coding — Can the AI work in a command-line terminal — running commands and finishing technical setup tasks the way a developer would? Higher is better. | — | 73% |
Humanity's Last Exam · no toolsMultidisciplinary reasoning — Humanity's Last Exam — extremely hard expert questions across many subjects, written so you can't just look up the answer. “No tools” means the AI answers on its own. Higher is better. | 21.4% | — |
GPQA DiamondScience — Graduate-level science questions in biology, physics, and chemistry — hard enough that subject-matter PhDs score around 65%. Higher is better. | 86% | — |
CharXiv ReasoningChart reasoning — Can the AI read and reason about complex charts and figures, not just text? Higher is better. | 78% | — |
MMMU-ProMultimodal reasoning — A tougher version of MMMU — college-level questions that mix images, diagrams, and text together. Higher is better. | 75.3% | — |
MMMUMultimodal — Tests the AI on understanding images and text together across many college subjects. Higher is better. | 81.7% | — |
threejsevalCommunity preference (Three.js) — Every model gets the same prompt — "the Eiffel Tower", "a glass fishbowl", "a robot arm picking toys into a box" — and builds a 3D scene in Three.js. Real people then see two scenes side by side, names hidden, and vote for the one they prefer. The votes become a chess-style Elo rating on threejseval.com, averaged across all the prompts. It measures whether the scene looks and moves right to a human eye, not whether the code passes a test. Higher is better. | — | 1137 |
| Overview | ||
| Company | Qwen | SpaceXAI |
| Release date | Apr 16 2026 | May 18 2026 |
| Access | Open Weight | Proprietary |
Other comparisons
Qwen3.6vsClaude Fable 5.1Composer 2.5vsClaude Fable 5.1Qwen3.6vsGPT-6 AstraComposer 2.5vsGPT-6 AstraQwen3.6vsGemini 3.8 FlashComposer 2.5vsGemini 3.8 FlashQwen3.6vsMuse Spark 1.3Composer 2.5vsMuse Spark 1.3Qwen3.6vsDeepSeek-V4.1-FlashComposer 2.5vsDeepSeek-V4.1-FlashQwen3.6vsMistral Medium 3.5Composer 2.5vsMistral Medium 3.5
Frequently asked questions
Composer 2.5 leads Qwen3.6 on 3 of the 3 benchmarks they both report (SWE-Bench Pro, SWE-Bench Multilingual, Terminal-Bench 2.0). Only Qwen3.6 has a verified first-party API price: $0.375 per million input tokens and $2.25 per million output tokens. No pay-as-you-go API rate is tracked for Composer 2.5. Qwen3.6 shipped 32 days before Composer 2.5, so benchmark comparisons should account for the intervening progress.
Qwen3.6 is open weight, while Composer 2.5 is proprietary.
On SWE-Bench Pro, Composer 2.5 leads at 54% vs Qwen3.6 at 49.5%. On SWE-Bench Multilingual, Composer 2.5 leads at 79.8% vs Qwen3.6 at 67.2%. On Terminal-Bench 2.0, Composer 2.5 leads at 69.3% vs Qwen3.6 at 51.5%.