Skip to content
Model Pareto

Qwen3.8 Flash-Next

Alibaba QwenOpen weightshuggingface.co ↗
Compare with nearest rivalsCompare with…
Input
$0.15/M tok
Output
$0.47/M tok
Blended
$0.23/M tok · 3:1
Speed
54tok/s
Context
262Ktokens

Open weights (125B total / 6B active); price is the hosted Qwen3.8-Flash.

Rankings

Self-hostingEstimated

How we estimate →

What it would cost to run these open weights yourself on rented GPUs, compared with the API price above.

Hardware
2× B200 180GB
FP8 weights
Throughput
~12,519 tok/s
many requests batched
Price
$0.098–0.87 /M tok
busy → light use
On your own machine
Mac Studio M5 Max 128GB (int4, ~140 tok/s single-stream)
single consumer GPU or Mac
Assumptions (5)
  • FP8 weights (180B params, 6B active per token) + 25% KV-cache headroom ≈ 225 GB
  • 2× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
  • ~12,519 output tok/s aggregate at batch 358 (bandwidth-bound); MoE compute scales with active params
  • Blended 3:1 input:output; prefill ~354,375 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
~#3 of 14 (estimated)~82.9%
  1. #1Claude Opus 596.0%
  2. #2Claude Sonnet 585.2%
  3. ~#3Qwen3.8 Flash-NextEstimated82.9%
  4. #3Gemini 3.1 Pro80.6%
No published score — estimated from related benchmarks
SWE-bench Pro (Public)Estimated
~#7 of 19 (estimated)~64.7%
  1. #1Claude Opus 5.589.9%
  2. #2Claude Sonnet 5.581.3%
  3. #6Hy4 preview65.7%
  4. ~#7Qwen3.8 Flash-NextEstimated64.7%
  5. #7GPT-5.6 Terra63.4%
No published score — estimated from related benchmarks
AA Coding Agent IndexEstimated
~#10 of 15 (estimated)~47.7
  1. #1Claude Opus 5.566.0
  2. #2Claude Fable 5.162.2
  3. #9Kimi K351.9
  4. ~#10Qwen3.8 Flash-NextEstimated47.7
  5. #10Grok 4.647.0
No published score — estimated from related benchmarks
Not comparable across labs (1)

Kept for reference, not counted in rankings: these use a lab-specific task set, answer key or scoring, so scores can't be compared fairly across labs. Only published scores are shown.

Writing & chat

MMMLUEstimated
~#2 of 5 (estimated)~91.5%
  1. #1Gemini 3.1 Pro92.6%
  2. ~#2Qwen3.8 Flash-NextEstimated91.5%
  3. #2Gemma 4 31B88.4%
No published score — estimated from related benchmarks
LMArena Text, Non-English (Elo)Estimated
~#15 of 54 (estimated)~1457 Elo
  1. #1Claude Opus 5.51501
  2. #2Claude Fable 5.1Few votes1496
  3. #14GPT-6 AstraFew votes1460
  4. ~#15Qwen3.8 Flash-NextEstimated1457
  5. #15GPT-5.6 Terra1456
No published score — estimated from related benchmarks
LMArena Text (Elo)Estimated
~#15 of 54 (estimated)~1468 Elo
  1. #1Claude Opus 5.51509
  2. #2Claude Fable 5.11501
  3. #14GLM-5.3 Flash1475
  4. ~#15Qwen3.8 Flash-NextEstimated1468
  5. #15GPT-5.6 Terra1466
No published score — estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)Estimated
~#9 of 26 (estimated)~1947 Elo
  1. #1GPT-6 Astra2173
  2. #2Claude Fable 5.12162
  3. #8Grok 4.72007
  4. ~#9Qwen3.8 Flash-NextEstimated1947
  5. #9Muse Spark 1.31906
No published score — estimated from related benchmarks

Vision

MMMUEstimated
~#2 of 3 (estimated)~71.3%
  1. #1Claude Haiku 4.573.2%
  2. ~#2Qwen3.8 Flash-NextEstimated71.3%
  3. #2Llama 4 Maverick69.4%
No published score — estimated from related benchmarks
OmniDocBench v1.5Estimated
~#4 of 8 (estimated)~88.9%
  1. #1MiniMax-M391.6%
  2. #2Kimi K391.1%
  3. #2Qwen3.8 27B91.1%
  4. ~#4Qwen3.8 Flash-NextEstimated88.9%
  5. #4Gemini 3.1 Pro85.3%
No published score — estimated from related benchmarks
ChartographyEstimated
~#10 of 23 (estimated)~27.9%
  1. #1GPT-6 Astra71.0%
  2. #2Claude Opus 5.566.3%
  3. #8Qwen3.8 Max (0902)29.1%
  4. ~#10Qwen3.8 Flash-NextEstimated27.9%
  5. #10Muse Spark 1.327.6%
No published score — estimated from related benchmarks

Hard reasoning

AIME (latest)Estimated
~#2 of 25 (estimated)~96.7%
  1. #1Inkling97.1%
  2. ~#2Qwen3.8 Flash-NextEstimated96.7%
  3. #2Muse Glimmer94.7%
No published score — estimated from related benchmarks

Agents

τ²-bench (Telecom)Estimated
~#2 of 116 (estimated)~98.3%
  1. #1Step 3.7 Flash98.5%
  2. ~#2Qwen3.8 Flash-NextEstimated98.3%
  3. #2Gemini 3.1 Pro95.6%
No published score — estimated from related benchmarks
OSWorld-VerifiedEstimated
~#3 of 12 (estimated)~83.6%
  1. #1Kimi K384.8%
  2. #2Qwen3.8 27B84.3%
  3. ~#3Qwen3.8 Flash-NextEstimated83.6%
  4. #3Nex-N2.5-Pro82.2%
No published score — estimated from related benchmarks
OSWorld 2.0Estimated
~#3 of 5 (estimated)~56.6%
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. ~#3Qwen3.8 Flash-NextEstimated56.6%
  4. #3GPT-5.6 Terra50.2%
No published score — estimated from related benchmarks
BrowseCompEstimated
~#5 of 18 (estimated)~90.4%
  1. #1Atria Dawn Preview92.5%
  2. #2GPT-6 Astra91.5%
  3. #4Claude Opus 590.8%
  4. ~#5Qwen3.8 Flash-NextEstimated90.4%
  5. #5Nex-N2.5-Pro89.7%
No published score — estimated from related benchmarks
Not comparable across labs (1)

Kept for reference, not counted in rankings: these use a lab-specific task set, answer key or scoring, so scores can't be compared fairly across labs. Only published scores are shown.