# LFM2.5-24B-A2B vs Step 5 Preview > Side-by-side benchmark comparison for LFM2.5-24B-A2B and Step 5 Preview across agentic, coding, multimodal, knowledge, reasoning, multilingual, and math tasks. - Canonical page: https://benchlm.ai/compare/lfm2-5-24b-a2b-vs-step-5-preview - Last updated: October 13, 2026 - Shared sourced benchmarks: 4 - HTML indexing: indexable - Ranking lane: BenchAlign v5.8 ## Quick Verdict Step 5 Preview leads by point estimate. Conditional score ranges do not establish rank confidence. Choose using the category evidence and your own workload tests. ## Summary - Step 5 Preview has the higher overall point estimate 68.98 to 37.8. Conditional score ranges do not establish rank confidence. - The clearest category separation is in agentic, where the averages are 30.2 for LFM2.5-24B-A2B and 63.7 for Step 5 Preview. - The biggest single benchmark swing is HLE in Knowledge, with scores of 7.4% and 46.5%. - LFM2.5-24B-A2B is the cheaper option on output tokens, which matters if you expect large responses or heavy interactive use. - Step 5 Preview also has the larger context window at 1M. ## Model Snapshot | Property | LFM2.5-24B-A2B | Step 5 Preview | |----------|----------|----------| | Creator | LiquidAI | StepFun | | Type | Open Weight | Pending | | Reasoning | Reasoning | Reasoning | | Context | 131K | 1M | | Overall Score | 37.8 | 68.98 | | Benchmarks Covered | 13 | 22 | | Pricing (input/output) | $0.10 / $1.00 | $1.00 / $2.70 | ## Category Breakdown ### Agentic - Winner: Directional only - LFM2.5-24B-A2B public-lane score: 30.2 (Estimated · #88/124) - Step 5 Preview public-lane score: 63.7 (Estimated · #17/124) | Benchmark | LFM2.5-24B-A2B | Step 5 Preview | Winner | |-----------|-----------|-----------|--------| | AutomationBench | 27.7% | 44.0% | Step 5 Preview | | BFCL v4 | 65.0% | Coming soon | Coming soon | | τ³-bench results | 8.0% | Coming soon | Coming soon | | PinchBench | 83.2% | Coming soon | Coming soon | | WebArena-Verified | 26.8% | Coming soon | Coming soon | | Terminal-Bench 2.1 | Coming soon | 85.0% | Coming soon | | Terminal-Bench 4.0 | Coming soon | 33.30% | Coming soon | | CyberGym | Coming soon | 84.7% | Coming soon | | Toolathlon-Verified | Coming soon | 74.1% | Coming soon | | MCP Atlas | Coming soon | 85.6% | Coming soon | | JobBench | Coming soon | 59.0% | Coming soon | | APEX-Agents | Coming soon | 37.8% | Coming soon | | DRACO | Coming soon | 83.3% | Coming soon | | BrowseComp | Coming soon | 88.7% | Coming soon | | Agents' Last Exam | Coming soon | 29.5% | Coming soon | ### Coding - Winner: Directional only - LFM2.5-24B-A2B public-lane score: 28.5 (Estimated · #98/149) - Step 5 Preview public-lane score: 60.6 (Estimated · #20/149) | Benchmark | LFM2.5-24B-A2B | Step 5 Preview | Winner | |-----------|-----------|-----------|--------| | SciCode | 32.7% | 58.9% | Step 5 Preview | | SWE-bench Verified | 51% | Coming soon | Coming soon | | DeepSWE | Coming soon | 67.7% | Coming soon | | Terminal-Bench 2.1 | Coming soon | 85.0% | Coming soon | | ProgramBench | Coming soon | 80.5% | Coming soon | | sweMarathon | Coming soon | 72.7% | Coming soon | | MLS-Bench Lite | Coming soon | 40.5% | Coming soon | ### Reasoning - Winner: Not comparable - LFM2.5-24B-A2B public-lane score: Coming soon - Step 5 Preview public-lane score: 83.4 (Unranked · 2 rankable rows) | Benchmark | LFM2.5-24B-A2B | Step 5 Preview | Winner | |-----------|-----------|-----------|--------| | CritPt | Coming soon | 20.9% | Coming soon | ### Multimodal & Grounded - Winner: Not comparable - LFM2.5-24B-A2B public-lane score: Coming soon - Step 5 Preview public-lane score: 74.6 (#24/54) | Benchmark | LFM2.5-24B-A2B | Step 5 Preview | Winner | |-----------|-----------|-----------|--------| | OfficeQA Pro | Coming soon | 60.3% | Coming soon | | MMMU-Pro | Coming soon | 76% | Coming soon | ### Knowledge - Winner: Directional only - LFM2.5-24B-A2B public-lane score: 40.8 (Estimated · #112/181) - Step 5 Preview public-lane score: 68.4 (Supported · #22/181) | Benchmark | LFM2.5-24B-A2B | Step 5 Preview | Winner | |-----------|-----------|-----------|--------| | GPQA | 74.4% | Coming soon | Coming soon | | GPQA-D | 74.4% | 93.5% | Step 5 Preview | | HLE | 7.4% | 46.5% | Step 5 Preview | ### Instruction Following - Winner: Not comparable - LFM2.5-24B-A2B public-lane score: 75.6 (#61/128) - Step 5 Preview public-lane score: Coming soon | Benchmark | LFM2.5-24B-A2B | Step 5 Preview | Winner | |-----------|-----------|-----------|--------| | IFEval | 96.8% | Coming soon | Coming soon | | IFBench | 74.7% | Coming soon | Coming soon | ### Mathematics - Winner: Not comparable - LFM2.5-24B-A2B public-lane score: Coming soon - Step 5 Preview public-lane score: Coming soon | Benchmark | LFM2.5-24B-A2B | Step 5 Preview | Winner | |-----------|-----------|-----------|--------| | MATH-500 | 96.0% | Coming soon | Coming soon | ## FAQ ### Which is better overall, LFM2.5-24B-A2B or Step 5 Preview? Step 5 Preview has the higher overall point estimate. Conditional score ranges do not establish rank confidence. ### Where is the biggest gap between LFM2.5-24B-A2B and Step 5 Preview? The widest category gap is in agentic, where the averages are 30.2 for LFM2.5-24B-A2B and 63.7 for Step 5 Preview. ### How many benchmarks does LFM2.5-24B-A2B cover on BenchLM? LFM2.5-24B-A2B currently has 13 sourced benchmark scores on BenchLM. ### How many benchmarks does Step 5 Preview cover on BenchLM? Step 5 Preview currently has 22 sourced benchmark scores on BenchLM. ## Related Comparisons - [LFM2.5-24B-A2B vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-lfm2-5-24b-a2b) - [Step 5 Preview vs Claude Opus 5.5](/compare/claude-opus-5-5-vs-step-5-preview) - [LFM2.5-24B-A2B vs GPT-6.1 Sol](/compare/gpt-6-1-sol-vs-lfm2-5-24b-a2b) - [Step 5 Preview vs GPT-6.1 Sol](/compare/gpt-6-1-sol-vs-step-5-preview) ## Explore More - [LFM2.5-24B-A2B profile](/models/lfm2-5-24b-a2b) - [Step 5 Preview profile](/models/step-5-preview) - [Compare Pricing](/llm-pricing) - [Alternative Finder](/tools/alternative-finder) - [LLM Selector](/tools/llm-selector) - [Overall Rankings](/best/overall)