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. | 8B | 30B |
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. | 256k | 200k |
| 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. | $0.15 | $0.00 |
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. | $0.15 | $0.00 |
Cached input priceA reduced rate for text you send over and over. If every request starts with the same instructions or the same document, the provider keeps a copy ready and charges less to read it again. | $0.015 | $0.00 |
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. | $0.15Mistral | — |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | $0.15Mistral | — |
| $37.50Mistral list price | $0.00Z.ai list price | |
| 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. | — | 59.2% |
LiveCodeBenchCompetitive coding — Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better. | — | 64% |
BrowseCompWeb browsing — Can the AI browse the web and track down hard-to-find answers? Higher is better. | — | 42.8% |
MATHReasoning — accuracy on the original competition mathematics test set | 87.6% | — |
AIME 2025Reasoning — single-answer accuracy on the 2025 invitational mathematics exam without tools | — | 91.6% |
| Overview | ||
| Company | Mistral | Z.ai |
| Release date | Dec 2 2025 | Jan 19 2026 |
| Access | Open Weight | Open Weight |
| Model details | View model | View model |
Other comparisons
Ministral 3 8B-25-12vsClaude Fable 5.2GLM-4.7-FlashvsClaude Fable 5.2Ministral 3 8B-25-12vsGPT-6.2 AstraGLM-4.7-FlashvsGPT-6.2 AstraMinistral 3 8B-25-12vsGemini 4 HydrogenGLM-4.7-FlashvsGemini 4 HydrogenMinistral 3 8B-25-12vsMuse EmberGLM-4.7-FlashvsMuse EmberMinistral 3 8B-25-12vsGrok 4.7GLM-4.7-FlashvsGrok 4.7Ministral 3 8B-25-12vsDeepSeek-V4.1-Flash-1017GLM-4.7-FlashvsDeepSeek-V4.1-Flash-1017Frequently asked questions
Ministral 3 8B-25-12 and GLM-4.7-Flash don't publish scores on any of the same benchmarks, so there's no direct head-to-head comparison. GLM-4.7-Flash is cheaper on both input and output: $0.00 vs $0.15 per million input tokens, and $0.00 vs $0.15 per million output tokens. Ministral 3 8B-25-12 shipped 48 days before GLM-4.7-Flash, so benchmark comparisons should account for the intervening progress.
Ministral 3 8B-25-12 has 8B parameters, while GLM-4.7-Flash has 30B. Context windows are 256k (Ministral 3 8B-25-12) vs 200k (GLM-4.7-Flash).
Direct benchmark comparisons are unavailable — Ministral 3 8B-25-12 and GLM-4.7-Flash don't publish scores on any of the same benchmarks.