Pairwise benchmark snapshot · Aug 16, 2026
In this head-to-head showdown, Llama 4 Scout delivers higher overall intelligence and human-preferred responses (Elo 1,412 vs 1,320), while Llama 4 Scout leads in SWE-bench software engineering benchmarks (52.4%), while Llama 4 Scout is more budget-friendly at $0.3/1M per 1M output tokens. Review the complete breakdown below to determine which model best fits your performance and budget requirements.
10 direct benchmark disciplines evaluated across capability, speed, and cost
Mistral code-specialist model.
Meta open-weight Llama 4 long-context sibling of Maverick. Common public API lists sit near $0.08–$0.30 / $0.30–$0.70 per 1M; we store a conservative hosted list until OpenRouter overwrites.
Currently comparing Codestral 25.01 vs Llama 4 Scout. Alternative family tiers available.
Δ 92 Arena Elo pts
155 tok/s
$0.3 / 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 | Llama 4 Scout | Advantage Delta |
|---|---|---|---|
| Preference Elo | +92 | ||
| Coding Elo | +4 | ||
| LiveBench | +5.6% | ||
| SWE-bench | +0.6% | ||
| GPQA Diamond | +12.8% | ||
| Time to first token | +20 ms | ||
| Output speed | 148 tok/s | 155 tok/s | +7 tok/s |
| Input price | $0.3/1M | $0.08/1M | +$0.22/1M |
| Output price | $0.9/1M | +$0.6/1M | |
| Context window | +9.7M |
Simulate monthly production API costs in USD (US Dollar).
Llama 4 Scout is estimated to save $16.70/month ($200/year).
Higher SWE-bench (52.4%).
Lower output list price ($0.3/1M).
Lower TTFT (130 ms).
Larger window (10M).
Llama 4 Scout is the side marked multimodal in the catalog.
| Target Workload | Recommended Pick | Evaluation Rationale |
|---|---|---|
| Repo / coding agents | Llama 4 Scout | Higher SWE-bench (52.4%). |
| High-volume chat | Llama 4 Scout | Lower output list price ($0.3/1M). |
| Voice / low-latency UI | Llama 4 Scout | Lower TTFT (130 ms). |
| Long-document RAG | Llama 4 Scout | Larger window (10M). |
| Screenshots / vision | Llama 4 Scout | Llama 4 Scout is the side marked multimodal in the catalog. |
$1/$5 hosted Claude vs a cheap long-context open sibling. The alternative-to-Sonnet pair that is honest.
Huge window, cheap hosted lists. The Haiku alternative that is actually open-weight.
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-llama-4-scout/opengraph-image
Verified Head-to-Head Telemetry
