Atlas

Benchmarks

LiveBench 2.0 (2026-11-06) Long Agentic

LiveBench 2.0 (2026-11-06 release) Long Agentic suite: Long-running agentic work where early decisions must hold up through later turns, simulated months or test cases (finance and operations apps, long-run operations, systems engineering, performance optimisation); 18 tasks, each scored 0-100. LiveBench 2.0's release of 2026-11-06 (Abacus AI) reruns Long Agentic, Coding, Research and Design on new tasks and keeps the Reasoning and Data Analysis runs and the October 25 Browser runs; each suite is one run per model at its highest supported effort, a task the run could not measure scores 0, and a model without an input a suite needs scores 0 there. A separate family from the 2026-10-14 release, whose Overall weights suites equally.

Models
36
Top score
82.9
Leader
Claude Fable 5.2
Updated
—
Resolution
—
Saturation
—
Signal
92ndpercentile
141 benchmarks from 0.8 to 2.9196 benchmarks from 2.9 to 5.0131 benchmarks from 5.0 to 7.1105 benchmarks from 7.1 to 9.282 benchmarks from 9.2 to 11.365 benchmarks from 11.3 to 13.327 benchmarks from 13.3 to 15.421 benchmarks from 15.4 to 17.58 benchmarks from 17.5 to 19.68 benchmarks from 19.6 to 21.76 benchmarks from 21.7 to 23.84 benchmarks from 23.8 to 25.95 benchmarks from 25.9 to 28.03 benchmarks from 28.0 to 30.12 benchmarks from 30.1 to 32.14 benchmarks from 32.1 to 34.21 benchmark from 34.2 to 36.33 benchmarks from 36.3 to 38.41 benchmark from 38.4 to 40.52 benchmarks from 40.5 to 42.61 benchmark from 42.6 to 44.71 benchmark from 44.7 to 46.81 benchmark from 48.9 to 50.9

66th strongest of 818

Signal-to-noise +16.23 for the LiveBench 2.0 (2026-11-06 release) family

Difficulty
Not estimated

Another benchmark represents this family in the fit.

Organizations
35 dated models · 6 frontier modelsClosed weightsOpen weights
Data: LiveBenchatlas.village.dev

LiveBench 2.0 (2026-11-06) Long Agentic leaderboard

#ModelSettingScoreSource
1Claude Fable 5.2x-high
82.9
2Gemini 4 Neonx-high
82.6
3GPT-6 Astramax
74.7
4Claude Fable 5.1max
74.2
5Claude Opus 5.5max
72.8
6Gemini 4 Carbonhigh
71.8
7GPT-6.2 Astramax
70.8
8GPT-6.1 Solmax
64.7
9GPT-6.2 Solmax
63.6
10Claude Sonnet 5.5max
62.5
11Gemini 4 Hydrogenx-high
61.8
12Gemini 4 Argonhigh
58.5
13GPT-6.1 Terramax
55.3
14Claude Haiku 5.5max
47.9
15Muse Spark 1.3max
47.9
16Qwen4-Maxx-high
46.6
17Muse Wondermax
46.2
18Grok 4.7x-high
45.3
19DeepSeek V4.1 Flash 1017max
38.8
20Kimi K3.1 Promax
38.7
21MiMo-V2.6-Prohigh
36.4
22Kimi K3max
36.0
23GPT-5.6 Terramax
25.9
24MiniMax M3.1 Flashmax
25.7
25GPT-6 Lunamax
24.7
26Mistral Large 4high
24.7
27GLM-5.3 Flashmax
20.7
28Step 5 Previewhigh
19.4
29Grok 4.6x-high
18.7
30Gemini 3.8 Flashhigh
16.5
31Qwen3.8 27Bx-high
15.9
32Muse Spark Openx-high
12.1
33Nemotron 3.5 Super VL 120B A12B BF16thinking
12.0
34Nemotron 3.5 Super VL 120B A12B NVFP4thinking
11.0
35Inklingnot reported
9.8
36Gemma 4 31B ITmax
8.5

Related benchmarks

Loading Atlas data…