Pairwise benchmark snapshot · Aug 1, 2026
In this head-to-head showdown, MiniMax M2.5 delivers higher overall intelligence and human-preferred responses (Elo 1,478 vs 1,320), while MiniMax M2.5 leads in SWE-bench software engineering benchmarks (75.8%), while Codestral 25.01 is more budget-friendly at $0.9/1M per 1M output tokens. Review the complete breakdown below to determine which model best fits your performance and budget requirements.
9 direct benchmark disciplines evaluated across capability, speed, and cost
Mistral code-specialist model.
MiniMax coding model. Tied near the top of official SWE-bench bash-only in Feb 2026.
Currently comparing Codestral 25.01 vs MiniMax M2.5. Alternative family tiers available.
Δ 158 Arena Elo pts
148 tok/s
$0.9 / 1M output
Relative percentile scores computed across all active models in the benchmark catalog.
Comparing top Western standard models with China's leading frontier rival across 6 skill dimensions. Tap any spoke or dot to inspect.
Category wins across reasoning intelligence, generation speed, and token cost.
Preference Elo, Coding proficiency, SWE-bench & LiveBench accuracy
Generation throughput and time to first token responsiveness
Cost per million tokens and max context window length
| Benchmark | Codestral 25.01 | MiniMax M2.5 | Advantage Delta |
|---|---|---|---|
| Preference Elo | +158 | ||
| Coding Elo | +134 | ||
| LiveBench | +10.4% | ||
| SWE-bench | +24% | ||
| GPQA Diamond | +17.8% | ||
| Time to first token | +100 ms | ||
| Output speed | 148 tok/s | 90 tok/s | +58 tok/s |
| Input price | $0.3/1M | $0.3/1M | — |
| Output price | $0.9/1M | $1.2/1M | +$0.3/1M |
| Context window | +51k |
Simulate monthly production API costs in USD (US Dollar).
Codestral 25.01 is estimated to save $4.50/month ($54/year).
Higher SWE-bench (75.8%).
Lower output list price ($0.9/1M).
Lower TTFT (150 ms).
Larger window (256k).
MiniMax M2.5 is the side marked multimodal in the catalog.
| Target Workload | Recommended Pick | Evaluation Rationale |
|---|---|---|
| Repo / coding agents | MiniMax M2.5 | Higher SWE-bench (75.8%). |
| High-volume chat | Codestral 25.01 | Lower output list price ($0.9/1M). |
| Voice / low-latency UI | Codestral 25.01 | Lower TTFT (150 ms). |
| Long-document RAG | Codestral 25.01 | Larger window (256k). |
| Screenshots / vision | MiniMax M2.5 | MiniMax M2.5 is the side marked multimodal in the catalog. |
Tied near the top of official SWE-bench bash-only in Feb 2026. Coding lists should still see it.
Jan 2025 code model. Still a valid cheap-coding compare. Not a chatbot.
No comments posted on this matchup yet. Be the first to share an evaluation note!
Dynamic OpenGraph banner generated at /compare/codestral-25-vs-minimax-m2-5/opengraph-image
Verified Head-to-Head Telemetry