Solar Open2 250B
- Input
- —
- Output
- —
- Speed
- ~91tok/sest.
- Context
- 1Mtokens
Long-tail import from Artificial Analysis (2026-09-24); AA variant: Solar Open2 250B. 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
- 4× B200 180GB
- FP8 weights
- Throughput
- ~21,909 tok/s
- many requests batched
- Price
- $0.12–1.11 /M tok
- busy → light use
- On your own machine
- Mac Studio M5 Ultra 512GB (int4, ~110 tok/s single-stream)
- single consumer GPU or Mac
Assumptions (5)
- FP8 weights (250B params, 15B active per token) + 40% KV-cache headroom ≈ 350 GB
- 4× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
- ~21,909 output tok/s aggregate at batch 627 (bandwidth-bound); MoE compute scales with active params
- Blended 3:1 input:output; prefill ~283,500 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
~#9 of 21 (estimated)~77.1%
- #1Claude Opus 596.0%
- #2Claude Sonnet 585.2%
- #7Mistral Medium 3.577.6%
- ~#9Solar Open2 250BEstimated77.1%
- #9Ripple-65B Preview76.5%
No published score — estimated from related benchmarks
SWE-bench Pro (Public)Estimated
~#17 of 24 (estimated)~57.6%
- #1Claude Fable 5.293.0%
- #2Claude Opus 5.589.9%
- #16Hy357.9%
- ~#17Solar Open2 250BEstimated57.6%
- #17Inkling54.3%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#37 of 46 (estimated)~51.1%
- #1Gemini 4 Argon78.8%
- #2DeepSeek V4.1 Flash 101777.0%
- #36Nex-N2.5-Pro55.8%
- ~#37Solar Open2 250BEstimated51.1%
- #37Qwen3.8 Max (0902)51.0%
No published score — estimated from related benchmarks
Terminal-Bench 4.0Estimated
~#33 of 102 (estimated)~16.6%
- #1Kimi K3.1 Pro67.2%
- #2Claude Sonnet 5.563.6%
- #32Grok 4.617.2%
- ~#33Solar Open2 250BEstimated16.6%
- #33GPT-5.514.6%
No published score — estimated from related benchmarks
LiveCodeBenchEstimated
~#15 of 28 (estimated)~75.0%
- #1Qwen3.8-Omni-Flash92.6%
- #2Qwen3.8 Flash-Next91.9%
- #14Maple-Preview75.1%
- ~#15Solar Open2 250BEstimated75.0%
- #15Mistral Small 464.0%
No published score — estimated from related benchmarks
SciCodeVerified
#50 of 11848.0%
- #1Claude Fable 5.267.7%
- #2Claude Opus 5.566.9%
- #49Quasar 438B48.1%
- #50Solar Open2 250B48.0%
- #50Solar Pro 448.0%
artificialanalysis.ai · 2026-09-24
AA Coding Agent IndexEstimated
~#23 of 26 (estimated)~37.1
- #1Claude Opus 5.561.7
- #2Claude Opus 559.7
- #22Claude Haiku 5.537.3
- ~#23Solar Open2 250BEstimated37.1
- #23Gemini 3.8 Flash33.3
No published score — estimated from related benchmarks
Hard reasoning
GPQA DiamondVerified
#53 of 19285.7%
- #1GPT-6 Astra96.1%
- #2Gemini 3.8 Flash95.3%
- #53Ring-2.6-1T85.7%
- #53Solar Open2 250B85.7%
- #58KAT Coder Pro V285.5%
artificialanalysis.ai · 2026-09-24
Humanity's Last ExamVerified
#60 of 19828.5%
- #1Claude Fable 5.274.3%
- #2Claude Opus 5.561.4%
- #59Qwen3.5 397B A17B29.0%
- #60Solar Open2 250B28.5%
- #61Nemotron 3 Ultra28.4%
artificialanalysis.ai · 2026-09-24
AIME (latest)Estimated
~#6 of 34 (estimated)~94.8%
- #1GLM-5.3 NVFP499.5%
- #2Beam 1.0 501B A23B97.7%
- #5Ripple-65B Preview95.0%
- ~#6Solar Open2 250BEstimated94.8%
- #6Muse Glimmer94.7%
No published score — estimated from related benchmarks
AA Intelligence IndexEstimated
~#41 of 103 (estimated)~27.4
- #1Gemini 4 Carbon59.4
- #2Claude Opus 5.557.6
- #40MiniMax-M329.2
- ~#41Solar Open2 250BEstimated27.4
- #41Quasar 438B26.7
No published score — estimated from related benchmarks
Agents
τ²-bench (Telecom)Estimated
~#28 of 120 (estimated)~83.3%
- #1Step 3.7 Flash98.5%
- #2DeepSeek-V4-Pro-0813-nvfp4-DSpark98.3%
- #26Nemotron 3 Ultra83.3%
- ~#28Solar Open2 250BEstimated83.3%
- #28Nex-N2-Pro81.6%
No published score — estimated from related benchmarks
OSWorld-VerifiedEstimated
~#11 of 15 (estimated)~68.7%
- #1Kimi K384.8%
- #2Qwen3.8 27B84.3%
- #10Nex-N2.5-Mini71.2%
- ~#11Solar Open2 250BEstimated68.7%
- #11Muse Glimmer65.9%
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)~69.5%
- #1Atria Dawn Preview92.5%
- #2GPT-6 Astra91.5%
- #18Nemotron 3.5 Bolt 74B-A9B70.0%
- ~#19Solar Open2 250BEstimated69.5%
- #19Mistral Medium 3.548.6%
No published score — estimated from related benchmarks
GDPval-AA (Elo)Estimated
~#29 of 92 (estimated)~1400 Elo
- #1Claude Opus 5.51866
- #2Muse Wonder Full1857
- #28Qwen3.8 27B1419
- ~#29Solar Open2 250BEstimated1400
- #29K2 Horizon 375B A23B1361
No published score — estimated from related benchmarks
AutomationBench-AAEstimated
~#36 of 85 (estimated)~25.8%
- #1Claude Fable 5.276.4%
- #2Muse Wonder Full71.3%
- #35K2 Horizon MoVA 36B A4B26.1%
- ~#36Solar Open2 250BEstimated25.8%
- #36Gemini 3.5 Flash-Lite25.0%
No published score — estimated from related benchmarks