MiMo-V2-Omni vs Devstral Small 2 vs Step 3.5 Flash 2603
MiMo-V2-Omni comes out ahead, 77 to 67 and 62 on our weighted score, though Devstral Small 2 is 14% cheaper per token.
- Our pick
XiaomiMiMo-V2-Omni
77/100- ECI—
- Price$0.14 / $0.28
- Context262K
Mistral AIDevstral Small 2
67/100- ECI—
- Price$0.10 / $0.30
- Context262K
StepFunStep 3.5 Flash 2603
62/100- ECI—
- Price$0.10 / $0.30
- Context256K
MiMo-V2-Omni is our pick
MiMo-V2-Omni is the better all-round choice, scoring 77/100 against Devstral Small 2 (67) and Step 3.5 Flash 2603 (62). It leads on inputs & features. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
- CapabilityNot enough dataNo independent benchmark covers every model here yet
- Lowest priceDevstral Small 2 and Step 3.5 Flash 2603Devstral Small 2 $0.15 · Step 3.5 Flash 2603 $0.15 · MiMo-V2-Omni $0.175 per 1M tokens (3:1 blend)
- Longest contextMiMo-V2-Omni and Devstral Small 2MiMo-V2-Omni 262,144 · Devstral Small 2 262,144 · Step 3.5 Flash 2603 256,000 tokens
- Widest inputsMiMo-V2-OmniMiMo-V2-Omni: Text, Images, PDFs, Audio, Video · Devstral Small 2: Text, Images · Step 3.5 Flash 2603: Text
- Self-hostingDevstral Small 2 and Step 3.5 Flash 2603Publishes downloadable weights (Apache-2.0)
| Measure | Weight | MiMo-V2-Omni | Devstral Small 2 | Step 3.5 Flash 2603 |
|---|---|---|---|---|
| Price | 50% | 86 | 89 | 89 |
| Inputs & features | 30% | 90 | 50 | 35 |
| Context window | 20% | 37 | 37 | 36 |
| Overall | 100% | 77/100 | 67/100 | 62/100 |
Left out because at least one model lacks the data: capability. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | MiMo-V2-OmniXiaomi | Devstral Small 2Mistral AI | Step 3.5 Flash 2603StepFun |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.14 | $0.10 (best) | $0.10 (best) |
| Output | $0.28 (best) | $0.30 | $0.30 |
| Cached input | $0.0028 (best) | — | $0.02 |
| Blended (3:1) | $0.175 | $0.15 (best) | $0.15 (best) |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official Xiaomi API | Median of 1 providers | Official StepFun (Global) API |
| Limits | |||
| Context window | 262,144 tokens (best) | 262,144 tokens (best) | 256,000 tokens |
| Max output | 131,072 tokens | 262,144 tokens (best) | 256,000 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | Yes | Yes | No |
| PDFs | Yes | No | No |
| Audio | Yes | No | No |
| Video | Yes | No | No |
| Reasoning | Yes | No | Yeslow · high |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | No | No |
| Availability | |||
| Weights | Proprietary | OpenApache-2.0 | Open |
| API model ID | mimo-v2-omni | — | step-3.5-flash-2603 |
| API providers | 2 | 1 | 3 (best) |
| Released | Mar 18, 2026 | Dec 9, 2025 | Apr 2, 2026 |
| Knowledge cutoff | Dec 2024 | Dec 2025 | Jan 2025 |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- MiMo-V2-Omni$1.96
- Devstral Small 2$1.60
- Step 3.5 Flash 2603$1.60
Which should you choose?
Which is better: MiMo-V2-Omni, Devstral Small 2 or Step 3.5 Flash 2603?
MiMo-V2-Omni is the better all-round choice, scoring 77/100 against Devstral Small 2 (67) and Step 3.5 Flash 2603 (62). It leads on inputs & features. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
Which is cheaper, MiMo-V2-Omni, Devstral Small 2 or Step 3.5 Flash 2603?
Devstral Small 2 is cheaper at $0.10 input / $0.30 output per million tokens (median across 1 API provider). Step 3.5 Flash 2603 costs $0.10 input / $0.30 output per million tokens (official StepFun (Global) API price); MiMo-V2-Omni costs $0.14 input / $0.28 output per million tokens (official Xiaomi API price). At a typical mix of three input tokens to one output token, that is $0.15 per million tokens for Devstral Small 2 versus $0.15 for Step 3.5 Flash 2603 (1× as much) and $0.175 for MiMo-V2-Omni (1.2× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. MiMo-V2-Omni has not been scored yet, Devstral Small 2 has not been scored yet and Step 3.5 Flash 2603 has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for MiMo-V2-Omni, Devstral Small 2 and Step 3.5 Flash 2603 yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. All three support tool calling for agent workflows.
Which has the bigger context window?
MiMo-V2-Omni and Devstral Small 2 have the largest context windows (262,144 and 262,144 tokens), against 256,000 for Step 3.5 Flash 2603. Maximum output per response: MiMo-V2-Omni up to 131,072, Devstral Small 2 up to 262,144, Step 3.5 Flash 2603 up to 256,000 tokens.
Which can read images, PDFs, audio or video?
MiMo-V2-Omni accepts text, images, PDFs, audio and video; Devstral Small 2 accepts text and images; Step 3.5 Flash 2603 accepts text. MiMo-V2-Omni handles the widest range of inputs.
Are any of these open source?
Devstral Small 2 and Step 3.5 Flash 2603 publishes its weights (Apache-2.0) and can be self-hosted; MiMo-V2-Omni is proprietary.
Which is newer?
Step 3.5 Flash 2603 is the newest, released Apr 2, 2026. MiMo-V2-Omni came out Mar 18, 2026; Devstral Small 2 came out Dec 9, 2025. Knowledge cutoff: MiMo-V2-Omni Dec 2024, Devstral Small 2 Dec 2025, Step 3.5 Flash 2603 Jan 2025.
How do you decide the winner?
Each model gets a 0–100 score on capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Oct 4, 2026.
