Instruction-Following Eval (IFEval)
We show this table for reference; we do not rank on it.
LFM2.5-24B-A2B has the highest reported IFEval score at 96.8%, ahead of LFM3-1.7B (96.7%) and Flint (95.1%) among 70 models. We compile the table from provider self-reports and secondary reports and keep it for display only; it does not affect overall rankings.
A benchmark of 541 prompts built from 25 verifiable instruction types. It tests whether a model follows checkable constraints such as keyword, length, casing, and response-format requirements.
Benchmark score on IFEval — November 13, 2026
We compile the IFEval rows from provider self-reports and secondary reports. LFM2.5-24B-A2B leads the table at 96.8%, followed by LFM3-1.7B (96.7%) and Flint (95.1%). We do not use these results to rank models overall.
LFM2.5-24B-A2B
LiquidAI
Reported by LiquidAIIFEval metric view not recorded
LFM3-1.7B
LiquidAI
Reported by LiquidAIIFEval metric view not recorded
Flint
Recursive
Reported by RecursiveIFEval metric view not recorded
70 modelsInstruction FollowingStaleDisplay onlyUpdated November 13, 2026
Benchmark score table (70 models)
None of the 70 results on this page has a documented run by the benchmark or an independent evaluator: 66 reported by the labs that built the models, 3 reported in another publisher’s document and 1 carried from an upstream page, where who ran it is not recorded.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | LFM2.5-24B-A2BLiquidAIReported by LiquidAIIFEval metric view not recorded | 96.8% | Reported by LiquidAI | Not reported | Open |
| 2 | LFM3-1.7BLiquidAIReported by LiquidAIIFEval metric view not recorded | 96.7% | Reported by LiquidAI | Not reported | Open |
| 3 | FlintRecursiveReported by RecursiveIFEval metric view not recorded | 95.1% | Reported by Recursive | Not reported | Open |
| 4 | Qwen3.5-27BAlibabaReported by AlibabaIFEval metric view not recorded | 95% | Reported by Alibaba | Not reported | Open |
| 5 | Agents-A1InternScienceReported by InternScienceIFEval metric view not recorded | 94.8% | Reported by InternScience | Not reported | Open |
| 6 | Agents-A1-4BInternScienceReported by InternScienceIFEval metric view not recorded | 94.8% | Reported by InternScience | Not reported | Open |
| 7 | Qwen3.7 PlusAlibabaReported by AlibabaIFEval metric view not recorded | 94.6% | Reported by Alibaba | Not reported | Closed |
| 8 | Qwen3.7 MaxAlibabaReported by AlibabaIFEval metric view not recorded | 94.3% | Reported by Alibaba | Not reported | Closed |
| 9 | Qwen3.6 PlusAlibabaReported by AlibabaIFEval metric view not recorded | 94.3% | Reported by Alibaba | Not reported | Closed |
| 10 | Kimi K2.5Moonshot AIAnother lab’s comparison tableIFEval metric view not recorded | 93.9% | Another lab’s comparison table | Not reported | Open |
| 11 | dots3-note PreviewDots StudioReported by Dots StudioIFEval metric view not recorded | 93.9% | Reported by Dots Studio | Not reported | Open |
| 12 | o3-miniOpenAIReported by OpenAIIFEval metric view not recorded | 93.9% | Reported by OpenAI | Not reported | Closed |
| 13 | Spark-X2.6-4BSparkLLMReported by SparkLLMIFEval metric view not recorded | 93.6% | Reported by SparkLLM | Not reported | Open |
| 14 | Qwen3.5-122B-A10BAlibabaReported by AlibabaIFEval metric view not recorded | 93.4% | Reported by Alibaba | Not reported | Open |
| 15 | Claude 3.7 Sonnet ThinkingAnthropicReported by AnthropicIFEval metric view not recorded | 93.2% | Reported by Anthropic | Not reported | Closed |
| 16 | GLM-5Z.AIAnother lab’s comparison tableIFEval metric view not recorded | 92.6% | Another lab’s comparison table | Not reported | Open |
| 17 | Qwen3.5 397BAlibabaReported by AlibabaIFEval metric view not recorded | 92.6% | Reported by Alibaba | Not reported | Open |
| 18 | K-EXAONE 2.0LG AI ResearchReported by LG AI ResearchIFEval metric view not recorded | 92.4% | Reported by LG AI Research | Not reported | Open |
| 19 | o1OpenAIReported by OpenAIIFEval metric view not recorded | 92.2% | Reported by OpenAI | Not reported | Closed |
| 20 | Llama 3.3 70BMetaReported by MetaIFEval metric view not recorded | 92.1% | Reported by Meta | Not reported | Open |
| 21 | Qwen3.5-35B-A3BAlibabaReported by AlibabaIFEval metric view not recorded | 91.9% | Reported by Alibaba | Not reported | Open |
| 22 | LFM2.5-8B-A1BLiquidAIReported by LiquidAIIFEval metric view not recorded | 91.8% | Reported by LiquidAI | Not reported | Open |
| 23 | Ternary Bonsai 2 27BPrism MLReported by Prism MLIFEval metric view not recorded | 91.3% | Reported by Prism ML | Not reported | Open |
| 24 | Claude Opus 4.5AnthropicAnother lab’s comparison tableIFEval metric view not recorded | 90.9% | Another lab’s comparison table | Not reported | Closed |
| 25 | Claude 3.7 SonnetAnthropicReported by AnthropicIFEval metric view not recorded | 90.8% | Reported by Anthropic | Not reported | Closed |
| 26 | Mellum2.1-12B-A2.5B-ThinkingJetBrainsReported by JetBrainsIFEval metric view not recorded | 90.6% | Reported by JetBrains | Not reported | Open |
| 27 | Kimi K2Moonshot AIReported by Moonshot AIIFEval metric view not recorded | 89.8% | Reported by Moonshot AI | Not reported | Closed |
| 28 | Nemotron Ultra 253BNVIDIAReported by NVIDIAIFEval metric view not recorded | 89.5% | Reported by NVIDIA | Not reported | Open |
| 29 | Qwen3 235B 2507AlibabaReported by AlibabaIFEval metric view not recorded | 88.7% | Reported by Alibaba | Not reported | Open |
| 30 | Llama 3.1 405BMetaReported by MetaIFEval metric view not recorded | 88.6% | Reported by Meta | Not reported | Open |
| 31 | GPT-4.1 miniOpenAIReported by OpenAIIFEval metric view not recorded | 88.5% | Reported by OpenAI | Not reported | Closed |
| 32 | Qwen3 235B 2507 (Reasoning)AlibabaReported by AlibabaIFEval metric view not recorded | 87.8% | Reported by Alibaba | Not reported | Open |
| 33 | Llama 3.1 70BMetaReported by MetaIFEval metric view not recorded | 87.5% | Reported by Meta | Not reported | Open |
| 34 | GPT-4.1OpenAIReported by OpenAIIFEval metric view not recorded | 87.4% | Reported by OpenAI | Not reported | Closed |
| 35 | MiniCPM5-2BOpenBMBReported by OpenBMBIFEval metric view not recorded | 86.7% | Reported by OpenBMB | Not reported | Open |
| 36 | DeepSeek V3DeepSeekCarried from an upstream pageIFEval metric view not recorded | 86.1% | Carried from an upstream page | Not reported | Open |
| 37 | Claude 3.5 HaikuAnthropicReported by AnthropicIFEval metric view not recorded | 85.9% | Reported by Anthropic | Not reported | Closed |
| 38 | ZAYA1-8BZyphraReported by ZyphraIFEval metric view not recorded | 85.6% | Reported by Zyphra | Not reported | Open |
| 39 | Qwen3 32B (Reasoning)AlibabaReported by AlibabaIFEval metric view not recorded | 85% | Reported by Alibaba | Not reported | Open |
| 40 | Phi-4-reasoning-plusMicrosoftReported by MicrosoftIFEval metric view not recorded | 84.9% | Reported by Microsoft | Not reported | Open |
| 41 | Sarvam 105BSarvamReported by SarvamIFEval metric view not recorded | 84.8% | Reported by Sarvam | Not reported | Open |
| 42 | QwQ-32BAlibabaReported by AlibabaIFEval metric view not recorded | 83.9% | Reported by Alibaba | Not reported | Open |
| 43 | Qwen3 235B-A22B (Reasoning)AlibabaReported by AlibabaIFEval metric view not recorded | 83.4% | Reported by Alibaba | Not reported | Open |
| 44 | Phi-4-reasoningMicrosoftReported by MicrosoftIFEval metric view not recorded | 83.4% | Reported by Microsoft | Not reported | Open |
| 45 | DeepSeek-R1DeepSeekReported by DeepSeekIFEval metric view not recorded | 83.3% | Reported by DeepSeek | Not reported | Open |
| 46 | Qwen3 235B-A22BAlibabaReported by AlibabaIFEval metric view not recorded | 83.2% | Reported by Alibaba | Not reported | Open |
| 47 | Qwen3 32BAlibabaReported by AlibabaIFEval metric view not recorded | 83.2% | Reported by Alibaba | Not reported | Open |
| 48 | GPT-4.1 nanoOpenAIReported by OpenAIIFEval metric view not recorded | 83.2% | Reported by OpenAI | Not reported | Closed |
| 49 | OrcaCyber-Nano-1.0OrcaRouterReported by OrcaRouterIFEval metric view not recorded | 82.8% | Reported by OrcaRouter | Not reported | Open |
| 50 | LFM2.5-VL-3BLiquidAIReported by LiquidAIIFEval metric view not recorded | 82.3% | Reported by LiquidAI | Not reported | Open |
| 51 | Ternary Bonsai 8BPrism MLReported by Prism MLIFEval metric view not recorded | 81.8% | Reported by Prism ML | Not reported | Open |
| 52 | Kanana-2 3B InstructKakaoReported by KakaoIFEval metric view not recorded | 81.0% | Reported by Kakao | Not reported | Open |
| 53 | Celeris-1CelerisReported by CelerisIFEval metric view not recorded | 80.8% | Reported by Celeris | Not reported | Closed |
| 54 | MiniCPM5-1BOpenBMBReported by OpenBMBIFEval metric view not recorded | 80.4% | Reported by OpenBMB | Not reported | Open |
| 55 | Llama 3.1 8BMetaReported by MetaIFEval metric view not recorded | 80.4% | Reported by Meta | Not reported | Open |
| 56 | 1-bit Bonsai 8BPrism MLReported by Prism MLIFEval metric view not recorded | 79.8% | Reported by Prism ML | Not reported | Open |
| 57 | Granite-4.0-H-1BIBMReported by IBMIFEval metric view not recorded | 78.5% | Reported by IBM | Not reported | Open |
| 58 | Kanana-2 1.3B InstructKakaoReported by KakaoIFEval metric view not recorded | 77.6% | Reported by Kakao | Not reported | Open |
| 59 | Granite-4.0-1BIBMReported by IBMIFEval metric view not recorded | 77.4% | Reported by IBM | Not reported | Open |
| 60 | LFM2.5-350MLiquidAIReported by LiquidAIIFEval metric view not recorded | 77.0% | Reported by LiquidAI | Not reported | Open |
| 61 | Mellum2-12B-A2.5B-ThinkingJetBrainsReported by JetBrainsIFEval metric view not recorded | 76.5% | Reported by JetBrains | Not reported | Open |
| 62 | Mellum2-12B-A2.5B-InstructJetBrainsReported by JetBrainsIFEval metric view not recorded | 75.8% | Reported by JetBrains | Not reported | Open |
| 63 | Ternary Bonsai 4BPrism MLReported by Prism MLIFEval metric view not recorded | 72.1% | Reported by Prism ML | Not reported | Open |
| 64 | LFM2.5-230MLiquidAIReported by LiquidAIIFEval metric view not recorded | 71.7% | Reported by LiquidAI | Not reported | Open |
| 65 | Ternary Bonsai 1.7BPrism MLReported by Prism MLIFEval metric view not recorded | 70.1% | Reported by Prism ML | Not reported | Open |
| 66 | 1-bit Bonsai 4BPrism MLReported by Prism MLIFEval metric view not recorded | 69.6% | Reported by Prism ML | Not reported | Open |
| 67 | 1-bit Bonsai 1.7BPrism MLReported by Prism MLIFEval metric view not recorded | 63% | Reported by Prism ML | Not reported | Open |
| 68 | Granite-4.0-H-350MIBMReported by IBMIFEval metric view not recorded | 61.6% | Reported by IBM | Not reported | Open |
| 69 | LFM2.5-VL-450MLiquidAIReported by LiquidAIIFEval metric view not recorded | 61.2% | Reported by LiquidAI | Not reported | Open |
| 70 | Granite-4.0-350MIBMReported by IBMIFEval metric view not recorded | 55.4% | Reported by IBM | Not reported | Open |
- Lab-reported
- Benchmark-run
- Independent
How to read this leaderboard
Editorial review by Glevd · 2026-07-15
Use IFEval to judge compliance with objectively checkable constraints after matching both scoring choices: prompt versus instruction level, and strict versus loose. An unlabeled IFEval number is incomplete because the four metrics answer different questions and can produce different scores.
Operator receipt: 70 sourced rows are currently displayable on this page; the leading published row is LFM2.5-24B-A2B at 96.8%.
Honest limit: IFEval covers constraints that can be verified automatically. It does not establish factual accuracy, usefulness, judgment, tone, or multi-turn reliability, and strict checks can reject a substantively compliant answer because of formatting. The current table marks public rows whose prompt/instruction and strict/loose metric view is not recorded; do not treat those values as a precise normalized race.
Among the reported IFEval rows, LFM2.5-24B-A2B is first at 96.8%. The third row is 1.7 points behind. The broader top-10 range is 2.8 points, so many of the published results sit in a relatively narrow band.
70 models have been evaluated on IFEval. The benchmark falls in the Instruction Following category. IFEval is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About IFEval
Year
2023
Tasks
541 prompts across 25 instruction types
Format
Constrained generation
Difficulty
Instruction precision
IFEval applies deterministic checkers and reports four views of the same responses: prompt-level strict, instruction-level strict, prompt-level loose, and instruction-level loose. Prompt-level scoring requires every instruction in a prompt to pass; loose scoring applies limited normalization before checking.
Freshness and provenance
Version
IFEval 2023
Refresh cadence
Static
Staleness state
Stale
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does IFEval measure?
IFEval measures whether a model follows checkable constraints in 541 prompts built from 25 instruction types. Examples include using required keywords, avoiding specified terms, meeting length limits, or producing a requested structure. Automatic checkers score each response, which makes the result reproducible for the constraints they can verify.
Which IFEval score should I compare?
Match all four metric choices before comparing models: prompt-level or instruction-level, and strict or loose. Prompt-level scoring requires every instruction in a prompt to pass. Loose scoring applies limited normalization to reduce formatting-related failures. An unlabeled IFEval number is not enough evidence for a precise comparison.
Does a high IFEval score mean a model is a good assistant?
No. A high IFEval score means the model obeyed more objectively checkable constraints under the selected metric. It does not show that the answer was factual, useful, well judged, stylistically appropriate, or reliable across a long conversation. Pair it with broader instruction tests and a workflow-specific prompt set.
Compare top models on IFEval
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.