# LFM3-1.7B vs LongCat-2.5

> Side-by-side benchmark comparison for LFM3-1.7B and LongCat-2.5 across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks.

- Canonical page: https://benchlm.ai/compare/lfm3-1-7b-vs-longcat-2-5
- Last updated: November 9, 2026

- Shared sourced benchmarks: 6
- HTML indexing: indexable
- Ranking lane: BenchAlign v5.8

## Quick Verdict

LFM3-1.7B leads by point estimate. Conditional score ranges do not establish rank confidence. Choose using the category evidence and your own workload tests.

## Summary

- LFM3-1.7B has the higher overall point estimate 21.36 to 0. Conditional score ranges do not establish rank confidence.
- The biggest single benchmark swing is Terminal-Bench 2.1 in Agentic, with scores of 12.9% and 88.4%.
- LFM3-1.7B is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use.
- LongCat-2.5 also has the larger context window at 1M.

## Model Snapshot

| Property | LFM3-1.7B | LongCat-2.5 |
|----------|----------|----------|
| Creator | LiquidAI | Meituan |
| Type | Open Weight | Open Weight |
| Reasoning | Reasoning | Hybrid |
| Context | 256K | 1M |
| Overall Score | 21.36 | null |
| Benchmarks Covered | 24 | 22 |
| Pricing (input/output) | $0.00 / $0.00 | $0.30 / $1.20 |

## Category Breakdown

### Agentic

- Winner: Not comparable
- LFM3-1.7B public-lane score: 16.1 (Estimated · #124/132)
- LongCat-2.5 public-lane score: Coming soon

| Benchmark | LFM3-1.7B | LongCat-2.5 | Winner |
|-----------|-----------|-----------|--------|
| AutomationBench | 41.3% | 51.8% | LongCat-2.5 |
| BFCL v4 | 71.4% | Coming soon | Coming soon |
| τ³-bench results | 68.0% | Coming soon | Coming soon |
| WebArena-Verified | 38.4% | Coming soon | Coming soon |
| MCP Atlas | 32.6% | 83.1% | LongCat-2.5 |
| ToolSandbox | 80.36% | Coming soon | Coming soon |
| Claw-Eval | 68.6% | 78.3% | LongCat-2.5 |
| HLE w/ tools | 13.9% | 63.1% | LongCat-2.5 |
| Terminal-Bench 2.1 | 12.9% | 88.4% | LongCat-2.5 |
| LiveClaw | 45.60% | Coming soon | Coming soon |
| Terminal-Bench 4.0 | Coming soon | 38.50% | Coming soon |
| BrowseComp | Coming soon | 91.9% | Coming soon |
| Agents' Last Exam | Coming soon | 38.5% | Coming soon |
| JobBench | Coming soon | 65.4% | Coming soon |
| skillsBench | Coming soon | 63.3% | Coming soon |
| SpreadsheetBench 2 | Coming soon | 34.9% | Coming soon |

### Coding

- Winner: Not comparable
- LFM3-1.7B public-lane score: 14.7 (Estimated · #152/158)
- LongCat-2.5 public-lane score: Coming soon

| Benchmark | LFM3-1.7B | LongCat-2.5 | Winner |
|-----------|-----------|-----------|--------|
| SciCode | 28.5% | Coming soon | Coming soon |
| SWE-bench Verified | 43.4% | Coming soon | Coming soon |
| SWE Multilingual | 23% | Coming soon | Coming soon |
| Terminal-Bench 2.1 | 12.9% | 88.4% | LongCat-2.5 |
| LiveCodeBench v6 | 73.5% | Coming soon | Coming soon |
| LiveCodeBench Pro | 31.7% | Coming soon | Coming soon |
| ProgramBench | Coming soon | 25.5% | Coming soon |
| DeepSWE | Coming soon | 64.6% | Coming soon |
| PaperBench | Coming soon | 82.1% | Coming soon |
| SWE-Atlas Codebase QnA | Coming soon | 58.3% | Coming soon |
| NL2Repo | Coming soon | 60.5% | Coming soon |

### Reasoning

- Winner: Not comparable
- LFM3-1.7B public-lane score: Coming soon
- LongCat-2.5 public-lane score: Coming soon

| Benchmark | LFM3-1.7B | LongCat-2.5 | Winner |
|-----------|-----------|-----------|--------|
| RULER (128K) | 65.56% | Coming soon | Coming soon |

### Multimodal & Grounded

- Winner: Not comparable
- LFM3-1.7B public-lane score: Coming soon
- LongCat-2.5 public-lane score: Coming soon

| Benchmark | LFM3-1.7B | LongCat-2.5 | Winner |
|-----------|-----------|-----------|--------|
| Vision2Web | Coming soon | 78.6% | Coming soon |
| CharXiv | Coming soon | 90.2% | Coming soon |
| MathVision w/ Python | Coming soon | 96.5% | Coming soon |
| Video-MME | Coming soon | 85.2% | Coming soon |

### Knowledge

- Winner: Not comparable
- LFM3-1.7B public-lane score: 24.1 (Estimated · #173/179)
- LongCat-2.5 public-lane score: Coming soon

| Benchmark | LFM3-1.7B | LongCat-2.5 | Winner |
|-----------|-----------|-----------|--------|
| GPQA | 63.4% | Coming soon | Coming soon |
| GPQA-D | 63.4% | Coming soon | Coming soon |
| HLE | 7.2% | Coming soon | Coming soon |
| FrontierScience | Coming soon | 77% | Coming soon |

### Instruction Following

- Winner: Not comparable
- LFM3-1.7B public-lane score: 71.7 (#76/157)
- LongCat-2.5 public-lane score: Coming soon

| Benchmark | LFM3-1.7B | LongCat-2.5 | Winner |
|-----------|-----------|-----------|--------|
| IFEval | 96.7% | Coming soon | Coming soon |
| IFBench | 73.0% | Coming soon | Coming soon |
| Multi-IF | 89.10% | Coming soon | Coming soon |
| IFStruct | 93.72% | Coming soon | Coming soon |

## FAQ

### Which is better overall, LFM3-1.7B or LongCat-2.5?

LFM3-1.7B has the higher overall point estimate. Conditional score ranges do not establish rank confidence.

### How many benchmarks does LFM3-1.7B cover on BenchLM?

LFM3-1.7B currently has 24 sourced benchmark scores on BenchLM.

### How many benchmarks does LongCat-2.5 cover on BenchLM?

LongCat-2.5 currently has 22 sourced benchmark scores on BenchLM.

## Related Comparisons

- [LFM3-1.7B vs LFM3-1.7B-Base](/compare/lfm3-1-7b-vs-lfm3-1-7b-base)
- [LongCat-2.5 vs LFM3-1.7B-Base](/compare/lfm3-1-7b-base-vs-longcat-2-5)
- [LFM3-1.7B vs LFM3-Audio-2B](/compare/lfm3-1-7b-vs-lfm3-audio-2b)
- [LongCat-2.5 vs LFM3-Audio-2B](/compare/lfm3-audio-2b-vs-longcat-2-5)

## Explore More

- [LFM3-1.7B profile](/models/lfm3-1-7b)
- [LongCat-2.5 profile](/models/longcat-2-5)
- [Compare Pricing](/llm-pricing)
- [Alternative Finder](/tools/alternative-finder)
- [LLM Selector](/tools/llm-selector)
- [Overall Rankings](/best/overall)