Compare AI models
| Specifications | ||
ParametersThe learned values that shape how a model responds. B means billion and T means trillion. For mixture-of-experts models, the total count is larger than the number active for each token. Parameter count alone does not measure quality. | — | 123B |
Context windowHow much information the model can work with in one request, measured in tokens — pieces of words or other data. This includes your question, attachments, conversation and reply. Limits can vary by model variant and provider. | 100k | 256k |
| API pricingUSD per 1M tokens · lower wins | ||
Cheapest inputLowest input rate across every provider on OpenRouter, the lab included. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | — | $0.40Mistral |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | — | $2.00Mistral |
| — | $180Mistral | |
| These models have no shared benchmark scores. | ||
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. | — | 72.2% |
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. | — | 61.3% |
HumanEvalFunction synthesis — 164 small Python problems: the AI is given a function's description and has to write the working function. This was the coding benchmark of the GPT-3.5 and GPT-4 era, before the field moved to fixing real bugs in real repositories. Higher is better. | 58.7% | — |
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. | — | 32.6% |
GSM8KGrade-school math — Grade-school maths word problems that take a few steps of arithmetic to work through. It separated the models of 2022 and 2023 sharply, then saturated. One caveat on the historical numbers: OpenAI included part of the GSM8K training set in GPT-4's pre-training mix, so GPT-4's score is not a clean few-shot result. Higher is better. | 86.7% | — |
| Overview | ||
| Company | Anthropic | Mistral |
| Release date | Aug 9 2023 | Dec 9 2025 |
| Access | Closed | Open Weight |
| Model details | View model | View model |
Other comparisons
Claude Instant 1.2vsGPT-6.1 SolDevstral 2 (123B)vsGPT-6.1 SolClaude Instant 1.2vsGemini 4 ArgonDevstral 2 (123B)vsGemini 4 ArgonClaude Instant 1.2vsMuse Spark OpenDevstral 2 (123B)vsMuse Spark OpenClaude Instant 1.2vsGrok 4.7Devstral 2 (123B)vsGrok 4.7Claude Instant 1.2vsDeepSeek-V4.1-Flash-1017Devstral 2 (123B)vsDeepSeek-V4.1-Flash-1017Claude Instant 1.2vsKimi K3Devstral 2 (123B)vsKimi K3Frequently asked questions
Claude Instant 1.2 and Devstral 2 (123B) don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. Claude Instant 1.2 shipped 853 days before Devstral 2 (123B), so benchmark comparisons should account for the intervening progress.
Context windows are 100k (Claude Instant 1.2) vs 256k (Devstral 2 (123B)). Claude Instant 1.2 is closed, while Devstral 2 (123B) is open weight.
Direct benchmark comparisons are unavailable — Claude Instant 1.2 and Devstral 2 (123B) don't publish scores on any of the same benchmarks.