Mistral: Mistral Small 3.2 24B
mistralai/mistral-small-3.2-24b-instruct
Cheapest provider
$0.07 / 1M
DeepInfra
Fastest provider (p95)
—
No throughput data yet — populated as traffic accumulates
Intelligence (composite)
64.6
MMLU-Pro · HumanEval · math · GPQA
Per-provider performance
Latency / throughput / uptime measured across providers over the last 30 minutes of live traffic. Atlas’s router weighs these per call (with the eval-gate signal) when picking a variant for Standard and Batch tiers.
| Provider | Quant | Input $/1M | Output $/1M | Latency p50 / p95 | Throughput p50 / p95 | Uptime 30m | Success |
|---|---|---|---|---|---|---|---|
| DeepInfra | q8· fp8 | $0.0750 | $0.2000 | — | — | 98.88% | — |
| Parasail | full· bf16 | $0.0900 | $0.6000 | — | — | 99.98% | — |
| Venice | q8· fp8 | $0.0938 | $0.2500 | — | — | 99.92% | — |
| Mistral | undisclosed | $0.1000 | $0.3000 | — | — | 99.91% | — |
“—” means live telemetry hasn’t accumulated enough recent traffic for that endpoint. “undisclosed” means the provider serves the model but doesn’t expose the quantization label (typically running fp8 / int8 internally).
Intelligence breakdown
Composite score is a weighted average of public benchmarks (30% MMLU-Pro, 25% code pass@1, 25% math, 20% GPQA). Numbers come from model cards and the Artificial Analysis intelligence harness; missing components are renormalised over what’s present.
MMLU-Pro
64.2
broad reasoning
Code
80.3
pass@1 (HumanEval / LiveCodeBench)
MATH
68.2
math accuracy
GPQA Diamond
41.2
hard reasoning
Source: Mistral Small 3.2 model card (2025)
How Atlas routes Mistral: Mistral Small 3.2 24B
- Realtime tier — direct passthrough at the upstream’s native precision. Best for hard latency / quality guarantees.
- Standard tier — Atlas picks the cheapest provider variant whose quantization has stayed green on your operation’s eval gates. For Mistral: Mistral Small 3.2 24B that’s currently DeepInfra at $0.07/1M.
- Batch tier — async, biggest discount. Roadmapped to use provider batch APIs (OpenAI / Anthropic) where available and queued spot capacity for open-weight workloads.