Across 2 comparable benchmarks, Step 3.5 Flash has better reported results on 2 and worse on 0.
Limited evidence2 of 2 shared results countedHow we compare
Devstral Medium is no longer available through its official API. API prices for this version are not shown.
Price basis: International rates vs listed rates. The price section below shows the applicable periods.
| Fact | Step 3.5 Flash | Devstral Medium |
|---|---|---|
| Context window | 256K tokens | 128K tokens |
| Input / 1M | $0.100 | — |
| Output / 1M | $0.300 | — |
| Blended (4:1) | $0.082 | — |
| Status | Available | Retired |
| Released | February 2, 2026 | July 10, 2025 |
Blended is 4:1 input to output with 90% cache hits at listed rates — set your own mix.
Mean of 2 shared benchmarks grouped by capability on a 0–100 scale. How averages are built
Step 3.5 Flash ahead on 2 of 2 counted
No shared comparable results for Abstract reasoning, Mathematics, Science & health, Computer & tool use, Research & search, Long context, Multimodal, Professional work & productivity, Cybersecurity.
Published scores, with matching modes where available. Gaps show Step 3.5 Flash relative to Devstral Medium. Reading the results
| Benchmark | Step 3.5 Flash | Devstral Medium | Gap |
|---|---|---|---|
| SWE-bench VerifiedRepository Engineering · Accuracy | 74.4Thinking · no tools | 61.6≠Normal · no tools | +12.8 pts#41 vs #77 of 108 |
| Terminal Bench HardAgentic Development · Accuracy | 27.3Thinking · tools | 9.1≠Normal · tools | +18.2 pts#75 vs #117 of 157 |
Repository Engineering · Accuracy
Agentic Development · Accuracy
A dash means no result is available.
| Benchmark | Step 3.5 Flash | Devstral Medium |
|---|---|---|
| AIME 2026 | 96.7 | — |
| AIME2025 | 99.8 | — |
| CritPt | 2.5 | — |
| IMO-AnswerBench | 86.7 | — |
| LiveCodeBench | 86.4 | — |
| Terminal Bench 2.0 | 51.0 | — |
| Claw Bench | 84.9 | — |
| Pinch Bench | 85.3 | — |
| PinchBench v2 | 79.3 | — |
| τ²-Bench | 88.2 | — |
| ARC-AGI-1 | 56.5 | — |
| BrowseComp | 69.0 | — |
Time-of-day providers use peak-hour rates; fixed-price providers use standard rates. Off-peak discounts remain in the published details.
Drag the handle above the bar for cache hits, the one below it for the input : output mix · 5% steps
Blended price = effective input rate at 90% cache hits × 80% + output rate × 20%
A blended cost needs a published input and output price on both sides, and one of these two does not have them yet.
Step 3.5 Flash: International pricing. StepFun official pricing (platform.stepfun.ai/docs/en/pricing/details) verified 2026-10-22; start date not published, verification date used
Devstral Medium: Listed text rates; pricing tier not confirmed.
| Price / 1M tokens | Step 3.5 Flash | Devstral Medium | Ratio |
|---|---|---|---|
| Input | $0.10 | — | — |
| Output | $0.30 | — | — |
| Cached input | $0.02 | — | — |
| Cache write | — | — | — |
| Blended | $0.082 | — | — |
Estimates use the selected token mix. Cache writes, tool fees and long-context surcharges are excluded.
Reported cost per task, paired with the score from the same run. Settings and sources may differ; costs do not affect the verdict.
| Benchmark | Step 3.5 Flash | Devstral Medium |
|---|---|---|
| AIME 2026 | Not reported |
| Attribute | Step 3.5 Flash | Devstral Medium |
|---|---|---|
| Availability | Available | Retired |
| Provider | StepFunAI | MistralAI |
| Model type | Chat model | Coding model |
| First release | February 2, 2026 | July 10, 2025 |
| Context window | 256K tokens | 128K tokens |
| Max output | 16K tokens | 4K tokens |
| Total parameters | 196B | — |
| Active parameters | 11B | — |
| Architecture | Mixture of Experts (MoE) | Undisclosed |
| Documented capabilities | Reasoning, Multilingual | Multilingual, Coding |
| Input modalities | Text | Text |
| Output modalities | Text | Text |
| Weights | Open weights | Closed / API only |
| Licence | Apache 2.0 | — |
| Commercial use | Commercial use allowed | Proprietary |
These are reported-score comparisons, not controlled head-to-head tests.
What people found comparing these models, or using either one. Comments also appear on the models' own pages.
Used Step 3.5 Flash or Devstral Medium? Sign in with Google to share what worked, what did not, and which API or setup you tested.
No community comments on either model yet.
Alternatives with the most benchmark overlap for this matchup.