Compare / head-to-head
GPT-5.4vs
Magistral Small 1.2
GPT-5.4 leads 1 of 1 shared benchmark. GPT-5.4 has the larger context window (1.1M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
1 – 0
GPT-5.4 leads
Cheaper per token
—
list price, input + output
Larger context
GPT-5.4
1.1M tokens
Providers
2 providers
OpenAI · Mistral AI
GPT-5.4
OpenAI · released 2026-03-05
textvision
Magistral Small 1.2
Mistral AI · released 2025-09-17
textvision
- Context
- 128K
- Max out
- —
- Input /1M
- —
- Output /1M
- —
- Cached /1M
- —
- Scores
- 4 · 3 core
Quality
Benchmark matrix
1 shared · 41 only GPT-5.4 · 3 only Magistral Small 1.2| Benchmark | GPT-5.4 | Magistral Small 1.2 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 1 | ||||
| GPQA Diamond | 92.8% ↗ | 70.07% ↗ | +22.7 pt | GPT-5.4 |
| Only GPT-5.4 reports · 41 | ||||
| AIME 2026 | 99.2% ↗matharena ⚠ | not reported | — | — |
| Humanity's Last Exam (no tools) | 39.8% ↗ | not reported | — | — |
| LMArena Elo | 1452.6 ↗ | not reported | — | — |
| APEX-Agents (Mercor) | 52.4% ↗ | not reported | — | — |
| BrowseComp | 82.7% ↗ | not reported | — | — |
| Chess Puzzles (Epoch AI run) | 44% ↗epoch run | not reported | — | — |
| DeepSWE v1.1 (Datacurve) | 51.8% ↗ | not reported | — | — |
| EBR-bench (Epoch AI run) | 25.4% ↗epoch run | not reported | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | 49% ↗epoch run | not reported | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 78.6% ↗epoch run | not reported | — | — |
| Furniture Assembly (Epoch AI run) | 37.5% ↗epoch run | not reported | — | — |
| GDPval (wins or ties) | 83% ↗ | not reported | — | — |
| GSO Opt@1 (GSO) | 30.4% ↗ | not reported | — | — |
| Humanity's Last Exam (with tools) | 52.1% ↗ | not reported | — | — |
| LiveBench Agentic Coding (LiveBench) | 53.8% ↗ | not reported | — | — |
| LiveBench Coding (LiveBench) | 77.5% ↗ | not reported | — | — |
| LiveBench Data Analysis (LiveBench) | 79.3% ↗ | not reported | — | — |
| LiveBench Instruction Following (LiveBench) | 70.2% ↗ | not reported | — | — |
| LiveBench Language (LiveBench) | 82.6% ↗ | not reported | — | — |
| LiveBench Mathematics (LiveBench) | 94.1% ↗ | not reported | — | — |
| LiveBench Reasoning (LiveBench) | 88.1% ↗ | not reported | — | — |
| LMArena Agent (LMArena) | 0.0105 ↗ | not reported | — | — |
| LMArena Vision (LMArena) | 1283.8 ↗ | not reported | — | — |
| LMArena WebDev (LMArena) | 1395.1 ↗ | not reported | — | — |
| MCP Atlas | 67.2% ↗ | not reported | — | — |
| MirrorCode (Epoch AI run) | 15.6% ↗epoch run | not reported | — | — |
| MMMU-Pro (no tools) | 81.2% ↗ | not reported | — | — |
| MMMU-Pro (with tools) | 82.1% ↗ | not reported | — | — |
| Mystery Game Puzzles (Epoch AI run) | 37% ↗epoch run | not reported | — | — |
| OSWorld-Verified | 75% ↗ | not reported | — | — |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 97.8% ↗epoch run | not reported | — | — |
| SAGE (Vals AI) | 43.3% ↗ | not reported | — | — |
| SimpleQA Verified | 45.1% ↗epoch run | not reported | — | — |
| SWE-Bench Pro | 57.7% ↗ | not reported | — | — |
| SWE-bench Verified (Epoch AI run) | 76.9% ↗epoch run | not reported | — | — |
| tau2-bench Banking Knowledge (Sierra) | 39.4% ↗ | not reported | — | — |
| Tau2-bench Telecom | 98.9% ↗ | not reported | — | — |
| Terminal-Bench 2.0 | 75.1% ↗ | not reported | — | — |
| Toolathlon | 54.6% ↗ | not reported | — | — |
| Vending-Bench 2 (Andon Labs) | 6144.18 ↗ | not reported | — | — |
| WeirdML (Håvard Tveit Ihle) | 77.7% ↗ | not reported | — | — |
| Only Magistral Small 1.2 reports · 3 | ||||
| AIME24 (pass@1) | not reported | 86.14% ↗ | — | — |
| AIME25 (pass@1) | not reported | 77.34% ↗ | — | — |
| LiveCodeBench (v5) | not reported | 70.88% ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is GPT-5.4 minus Magistral Small 1.2 in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | GPT-5.4 | Magistral Small 1.2 | Edge |
|---|---|---|---|
| Context window | 1.1M tokens | 128K tokens | GPT-5.4 |
| Max output | 128K tokens | — | — |
| Input price / 1M | $2.5 ↗ | — | — |
| Output price / 1M | $15 ↗ | — | — |
| Cached input / 1M | $0.25 ↗ | — | — |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $17.5 | — | — |
| Modalities | text · vision | text · vision | Tie |
| Released | 2026-03-05 | 2025-09-17 | — |
| Cited benchmark scores | 50 | 4 | — |
Reliability
Provider status
Live from /statusMore matchups
GPT-5.4 vs …
Models sharing the most benchmarksMore matchups
Magistral Small 1.2 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run GPT-5.4 and Magistral Small 1.2 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.