# Inkling vs Step 3.7 Flash

> Side-by-side benchmark comparison for Inkling and Step 3.7 Flash across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/inkling-vs-step-3-7-flash
- Last updated: October 14, 2026

- Shared sourced benchmarks: 4
- HTML indexing: indexable
- Ranking lane: BenchAlign v5.8

## Quick Verdict

Inkling leads by point estimate. Conditional score ranges do not establish rank confidence. Choose using the category evidence and your own workload tests.

## Summary

- Inkling has the higher overall point estimate 52.13 to 43.49. Conditional score ranges do not establish rank confidence.
- The clearest category separation is in knowledge, where the averages are 54.5 for Inkling and 45 for Step 3.7 Flash.
- The biggest single benchmark swing is Terminal-Bench 2.1 in Agentic, with scores of 63.8% and 59.5%.
- Step 3.7 Flash is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.
- Inkling also has the larger context window at 1M.

## Model Snapshot

| Property | Inkling | Step 3.7 Flash |
|----------|----------|----------|
| Creator | Thinking Machines Lab | StepFun |
| Type | Open Weight | Open Weight |
| Reasoning | Hybrid | Reasoning |
| Context | 1M | 256K |
| Overall Score | 52.13 | 43.49 |
| Benchmarks Covered | 24 | 11 |
| Pricing (input/output) | $1.87 / $4.68 | $0.20 / $1.15 |

## Category Breakdown

### Agentic

- Winner: Directional only
- Inkling public-lane score: 36.2 (Supported · #72/124)
- Step 3.7 Flash public-lane score: 36.1 (Estimated · #73/124)

| Benchmark | Inkling | Step 3.7 Flash | Winner |
|-----------|-----------|-----------|--------|
| Terminal-Bench 2.1 | 63.8% | 59.5% | Inkling |
| BrowseComp | 77.1% | 75.8% | Inkling |
| MCP Atlas | 74.1% | Coming soon | Coming soon |
| Terminal-Bench 2.1 (Vals) | 47.6% | Coming soon | Coming soon |
| CWE-bench v1 | 37.0% | Coming soon | Coming soon |
| DeepSearchQA | Coming soon | 92.8% | Coming soon |
| Toolathlon | Coming soon | 49.5% | Coming soon |
| Claw-Eval | Coming soon | 67.1% | Coming soon |
| HLE w/ tools | Coming soon | 47.2% | Coming soon |
| Gert Labs | Coming soon | 51.57% | Coming soon |

### Coding

- Winner: Directional only
- Inkling public-lane score: 34 (Supported · #80/149)
- Step 3.7 Flash public-lane score: 30.5 (Estimated · #94/149)

| Benchmark | Inkling | Step 3.7 Flash | Winner |
|-----------|-----------|-----------|--------|
| SWE-bench Verified | 77.6% | Coming soon | Coming soon |
| SWE-bench Pro | 54.3% | 56.3% | Step 3.7 Flash |
| Terminal-Bench 2.1 | 63.8% | 59.5% | Inkling |
| FrontierSWE v2 | 4.1% | Coming soon | Coming soon |
| LiveCodeBench (Vals) | 85.5% | Coming soon | Coming soon |
| SWE-bench (Vals) | 77.6% | Coming soon | Coming soon |

### Reasoning

- Winner: Not comparable
- Inkling public-lane score: 79 (#14/29)
- Step 3.7 Flash public-lane score: 76.5 (Unranked · 2 rankable rows)

### Multimodal & Grounded

- Winner: Not comparable
- Inkling public-lane score: 67.5 (#35/54)
- Step 3.7 Flash public-lane score: 75.4 (Unranked · 3 rankable rows)

| Benchmark | Inkling | Step 3.7 Flash | Winner |
|-----------|-----------|-----------|--------|
| MMMU-Pro | 73.5% | Coming soon | Coming soon |
| CharXiv | 82% | Coming soon | Coming soon |
| CharXiv w/o tools | 78.1% | Coming soon | Coming soon |
| SimpleVQA | Coming soon | 79.2% | Coming soon |
| V* | Coming soon | 95.3% | Coming soon |

### Knowledge

- Winner: Directional only
- Inkling public-lane score: 54.5 (Supported · #62/181)
- Step 3.7 Flash public-lane score: 45 (Estimated · #95/181)

| Benchmark | Inkling | Step 3.7 Flash | Winner |
|-----------|-----------|-----------|--------|
| GPQA | 87.9% | Coming soon | Coming soon |
| GPQA-D | 87.9% | Coming soon | Coming soon |
| HLE | 46% | Coming soon | Coming soon |
| HLE w/o tools | 30% | Coming soon | Coming soon |
| GPQA Diamond (Vals) | 87.1% | Coming soon | Coming soon |
| MMLU-Pro (Vals) | 86.3% | Coming soon | Coming soon |

### Instruction Following

- Winner: Directional only
- Inkling public-lane score: 87.6 (#35/128)
- Step 3.7 Flash public-lane score: 82.1 (#53/128)

| Benchmark | Inkling | Step 3.7 Flash | Winner |
|-----------|-----------|-----------|--------|
| IFBench | 79.8% | Coming soon | Coming soon |

### Mathematics

- Winner: Not comparable
- Inkling public-lane score: 58.3 (#11/12)
- Step 3.7 Flash public-lane score: Coming soon

| Benchmark | Inkling | Step 3.7 Flash | Winner |
|-----------|-----------|-----------|--------|
| AIME26 | 97.1% | Coming soon | Coming soon |
| FrontierMath v2 (Tiers 1-3) | 33.333% | Coming soon | Coming soon |
| FrontierMath v2 (Tier 4) | 4.878% | Coming soon | Coming soon |

## FAQ

### Which is better overall, Inkling or Step 3.7 Flash?

Inkling has the higher overall point estimate. Conditional score ranges do not establish rank confidence.

### Where is the biggest gap between Inkling and Step 3.7 Flash?

The widest category gap is in knowledge, where the averages are 54.5 for Inkling and 45 for Step 3.7 Flash.

### How many benchmarks does Inkling cover on BenchLM?

Inkling currently has 24 sourced benchmark scores on BenchLM.

### How many benchmarks does Step 3.7 Flash cover on BenchLM?

Step 3.7 Flash currently has 11 sourced benchmark scores on BenchLM.

## Related Comparisons

- [Inkling vs Inkling-Small](/compare/inkling-vs-inkling-small)
- [Step 3.7 Flash vs Inkling-Small](/compare/inkling-small-vs-step-3-7-flash)
- [Inkling vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-inkling)
- [Step 3.7 Flash vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-step-3-7-flash)

## Explore More

- [Inkling profile](/models/inkling)
- [Step 3.7 Flash profile](/models/step-3-7-flash)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)