Inkling 2 Small vs MiMo V2.6 Flash
Both models appear in the same evaluation table on the official model card, which is the only place they are measured under identical conditions and reported by the same party.
Across 34 benchmarks where both models report a score, Inkling 2 Small is higher on 28 and MiMo V2.6 Flash is higher on 6 . Inkling 2 Small has 594B total parameters with 17B active, and is priced at $1 per million output tokens on Tinker. Its largest margin is on AA-Briefcase v1.1. Its largest deficit is on Code2Video-AA.
Head to head
- Position
- Inkling 2 Small: the subject of this site · MiMo V2.6 Flash: open weights, smaller class
- Bechmarks scored
- 34 shared rows, out of 71 total
- Record
- 28 wins · 6 losses · 0 ties
Largest differences
| Benchmark | Inkling 2 Small | MiMo V2.6 Flash | Gap |
|---|---|---|---|
| AA-Briefcase v1.1 · Knowledge Work | 1,527 | 1,487 | + 40 |
| Code2Video-AA · Visual Creation | 1,020 | 1,052 | -32 |
| Chartography · Visual Reasoning | 36.7% | 7.5% | + 29.2 |
| AA-Omniscience v1.1 · Factuality | 16.1 | -10.8 | + 26.9 |
| CritPt v2 · STEM Reasoning | 55.7% | 29.5% | + 26.2 |
| MCP Atlas · Search and Tools | 85.4% | 62.9% | + 22.5 |
| ComplexConstraints v1.0 · Instruction Following | 44.5% | 22.1% | + 22.4 |
| Charxiv RQ · Visual Reasoning | 85.6% | 66.2% | + 19.4 |
| GDP.pdf · Visual Reasoning | 25.2% | 9% | + 16.2 |
| GDPVal-AA v2.1 · Knowledge Work | 1,630 | 1,614 | + 16 |
Every shared benchmark
| Benchmark | Inkling 2 Small | MiMo V2.6 Flash | Leader |
|---|---|---|---|
| Third Party Index | |||
| AA Index (v5.0) | 30.6 | 26.7 | Inkling 2 Small |
| Vals Index (v2.2) | 56.1% | 54.8% | Inkling 2 Small |
| Coding | |||
| FrontierCode v2.0 | 25.4% | 21.6% | Inkling 2 Small |
| DeepSWE v1.1 | 72.1% | 67.9% | Inkling 2 Small |
| Terminal Bench Science v0.1 | 5.7% | 4.3% | Inkling 2 Small |
| Terminal Bench v4.0 | 13.1% | 22.7% | MiMo V2.6 Flash |
| Terminal Bench 2.1 · Inkling 2 Small: Pi harness | 85.1% | 87.6% | MiMo V2.6 Flash |
| Vibe Code Bench v1.1 | 81.6% | 79% | Inkling 2 Small |
| Autoresearch | |||
| BioMysteryBench | 70% | 69.3% | Inkling 2 Small |
| Search and Tools | |||
| MCP Atlas · Full / Public | 85.4% | 62.9% | Inkling 2 Small |
| Knowledge Work | |||
| AA-Briefcase v1.1 | 1,527 | 1,487 | Inkling 2 Small |
| GDPVal-AA v2.1 | 1,630 | 1,614 | Inkling 2 Small |
| JobBench | 70.5% | 61.2% | Inkling 2 Small |
| Enterprise Operations | |||
| AutomationBench v1.0.6 | 45.6% | 52.3% | MiMo V2.6 Flash |
| Finance Work | |||
| Finance Agent v2.1 | 67.5% | 63.3% | Inkling 2 Small |
| Excel Modeling Benchmark v1.1 | 69.5% | 68.9% | Inkling 2 Small |
| Tax Agent Bench | 65.8% | 59.9% | Inkling 2 Small |
| Legal Work | |||
| Harvey LAB v1.1 | 21.7% | 13.3% | Inkling 2 Small |
| Legal Research Bench | 46.6% | 38% | Inkling 2 Small |
| Visual Creation | |||
| Code Arena | 1,653 | 1,637 | Inkling 2 Small |
| Code2Video-AA | 1,020 | 1,052 | MiMo V2.6 Flash |
| Visual Perception | |||
| Humanity’s Sixth Sense | 33.3% | 25.4% | Inkling 2 Small |
| Visual Reasoning | |||
| GDP.pdf | 25.2% | 9% | Inkling 2 Small |
| Charxiv RQ · original / with Python | 85.6% | 66.2% | Inkling 2 Small |
| Chartography · original / with Python | 36.7% | 7.5% | Inkling 2 Small |
| SAGE v1.1 | 88.6% | 83.3% | Inkling 2 Small |
| STEM Reasoning | |||
| CritPt v2 · Full / Hard | 55.7% | 29.5% | Inkling 2 Small |
| HLE-Diamond | 28.7% | 13.7% | Inkling 2 Small |
| HLE (text only) | 44.1% | 35.1% | Inkling 2 Small |
| Long Context | |||
| LongBench v2 | 70.6% | 60.7% | Inkling 2 Small |
| Factuality | |||
| AA-Omniscience v1.1 · score | 16.1 | -10.8 | Inkling 2 Small |
| Instruction Following | |||
| ComplexConstraints v1.0 | 44.5% | 22.1% | Inkling 2 Small |
| Safety | |||
| StrongREJECT (none) | 97.1% | 97.8% | MiMo V2.6 Flash |
| Cybersecurity | |||
| CyberBench v1.1 | 72.7% | 75.4% | MiMo V2.6 Flash |
Reading the comparison honestly
These numbers come from one table published by Thinking Machines, using third-party evaluations where those exist. That makes them comparable to each other, but it does not make them neutral: the vendor chose which benchmarks to run and how to present them, and a benchmark that a model loses has a way of not appearing.
The models are also not the same shape of thing. MiMo V2.6 Flash sits in the open weights, smaller class tier of the card's own grouping, while Inkling 2 Small is the subject of this site. A win against a smaller model is less meaningful than a win against a larger one, and the table does not adjust for that.