LFM2.5-2.6B
- Input
- —
- Output
- —
- Speed
- ~110tok/sest.
- Context
- 128Ktokens
Long-tail import from Artificial Analysis (2026-09-24); AA variant: LFM2.5-2.6B. Price = AA median host price.
Rankings
Self-hostingEstimated
How we estimate →What it would cost to run these open weights yourself on rented GPUs. No API sells this model, so this is its price on the chart.
- Hardware
- 1× B200 180GB
- BF16 weights
- Throughput
- ~19,533 tok/s
- many requests batched
- Price
- $0.037–0.33 /M tok
- busy → light use
- On your own machine
- RTX 5090 32GB (int4, 300+ tok/s single-stream)
- single consumer GPU or Mac
Assumptions (5)
- BF16 weights (2.69B params) + 25% KV-cache headroom ≈ 7 GB
- 1× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
- ~19,533 output tok/s aggregate at batch 256 (bandwidth-bound)
- Blended 3:1 input:output; prefill ~197,607 tok/s
- Low = 75% utilization at the low GPU price; high = 20% at the high price
Benchmark scores
~ italic, dashed = no published score yet, estimated from related benchmarks.
Coding
SWE-bench VerifiedEstimated
~last of 21 (estimated)below measured range (<51.0%)
No published score — estimated from related benchmarks
SWE-bench Pro (Public)Estimated
~last of 24 (estimated)below measured range (<43.8%)
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#45 of 46 (estimated)~18.6%
- #1Gemini 4 Argon78.8%
- #2DeepSeek V4.1 Flash 101777.0%
- #44Nemotron 3.5 Super VL24.6%
- ~#45LFM2.5-2.6BEstimated18.6%
- #45Nemotron 3.5 Super VL NVFP418.2%
No published score — estimated from related benchmarks
Terminal-Bench 4.0Estimated
~#63 of 102 (estimated)~0.1%
- #1Kimi K3.1 Pro67.2%
- #2Claude Sonnet 5.563.6%
- #53Trinity Large Thinking0.5%
- ~#63LFM2.5-2.6BEstimated0.1%
- #63A.X-K20.0%
No published score — estimated from related benchmarks
LiveCodeBenchEstimated
~#25 of 28 (estimated)~32.4%
- #1Qwen3.8-Omni-Flash92.6%
- #2Qwen3.8 Flash-Next91.9%
- #24Devstral Small 234.8%
- ~#25LFM2.5-2.6BEstimated32.4%
- #25Ministral 3 8B30.3%
No published score — estimated from related benchmarks
SciCodeVerified
#117 of 11814.4%
- #1Claude Fable 5.267.7%
- #2Claude Opus 5.566.9%
- #116Ministral 3 3B15.3%
- #117LFM2.5-2.6B14.4%
- #118K2 Horizon 0.9B6.9%
artificialanalysis.ai · 2026-09-24
AA Coding Agent IndexEstimated
~last of 26 (estimated)below measured range (<26.3)
No published score — estimated from related benchmarks
Hard reasoning
GPQA DiamondVerified
#145 of 19255.8%
- #1GPT-6 Astra96.1%
- #2Gemini 3.8 Flash95.3%
- #144Granite 4.2 3B55.9%
- #145LFM2.5-2.6B55.8%
- #146NVIDIA Nemotron Nano 9B V255.7%
artificialanalysis.ai · 2026-09-24
Humanity's Last ExamVerified
#143 of 1986.2%
- #1Claude Fable 5.274.3%
- #2Claude Opus 5.561.4%
- #143LFM2.5-1.2B-Thinking6.2%
- #143LFM2.5-2.6B6.2%
- #145Olmo 3 7B Think6.0%
artificialanalysis.ai · 2026-09-24
AIME (latest)Estimated
~#23 of 34 (estimated)~42.9%
- #1GLM-5.3 NVFP499.5%
- #2Beam 1.0 501B A23B97.7%
- #22Phi-4-reasoning-plus78.0%
- ~#23LFM2.5-2.6BEstimated42.9%
- #23Mistral Large 338.0%
No published score — estimated from related benchmarks
AA Intelligence IndexEstimated
~#98 of 103 (estimated)~6.0
- #1Gemini 4 Carbon59.4
- #2Claude Opus 5.557.6
- #97Celeris-16.3
- ~#98LFM2.5-2.6BEstimated6.0
- #98Ministral 3 14B6.0
No published score — estimated from related benchmarks
Agents
τ²-bench (Telecom)Estimated
~#58 of 120 (estimated)~35.2%
- #1Step 3.7 Flash98.5%
- #2DeepSeek-V4-Pro-0813-nvfp4-DSpark98.3%
- #57North Mini Code37.4%
- ~#58LFM2.5-2.6BEstimated35.2%
- #58Sarvam 30B34.5%
No published score — estimated from related benchmarks
OSWorld-VerifiedEstimated
~#12 of 15 (estimated)~56.5%
- #1Kimi K384.8%
- #2Qwen3.8 27B84.3%
- #11Muse Glimmer65.9%
- ~#12LFM2.5-2.6BEstimated56.5%
- #12Nemotron 3.5 Super VL NVFP452.3%
No published score — estimated from related benchmarks
OSWorld 2.0Estimated
~last of 5 (estimated)below measured range (<47.5%)
No published score — estimated from related benchmarks
BrowseCompEstimated
~#19 of 22 (estimated)~50.3%
- #1Atria Dawn Preview92.5%
- #2GPT-6 Astra91.5%
- #18Nemotron 3.5 Bolt 74B-A9B70.0%
- ~#19LFM2.5-2.6BEstimated50.3%
- #19Mistral Medium 3.548.6%
No published score — estimated from related benchmarks
GDPval-AA (Elo)Estimated
~#66 of 92 (estimated)~536 Elo
- #1Claude Opus 5.51866
- #2Muse Wonder Full1857
- #65Gemma 4 26B A4B553
- ~#66LFM2.5-2.6BEstimated536
- #66Devstral 2531
No published score — estimated from related benchmarks
AutomationBench-AAEstimated
~#73 of 85 (estimated)~0.8%
- #1Claude Fable 5.276.4%
- #2Muse Wonder Full71.3%
- #72Nemotron 3.5 Lightning0.8%
- ~#73LFM2.5-2.6BEstimated0.8%
- #73Ministral 3 3B0.7%
No published score — estimated from related benchmarks