HolyGrAIL

All tasks / Models

Inkling

Thinking Machines Lab's open-weights multimodal MoE (Apache-2.0) with 975B total and 41B active parameters.

  • Thinking Machines Lab
  • Open weights
  • Released 14 Jul 2026

Official page

Where it ranks

  • A
    Chat & assistants
    Models · Local · 1,441 on LMArena Text Arena (overall, style control)
  • B
    Writing
    Models · Local · 1,381 on LMArena Text Arena: Creative Writing (style control)
  • C
    Coding
    Models · Local · 1,414 on LMArena WebDev / Code Arena (overall)
  • B
    Image understanding
    Models · Local · 73.5% on Artificial Analysis MMMU-Pro (independent evaluation) (secondary benchmark)
  • C
    Vector graphics & logos
    Models · Local · 1,111 on Design Arena SVG
  • C
    Workflow automation
    Models · Local · 5% on AutomationBench-AA (Artificial Analysis)

Facts

Vendor
Thinking Machines Lab
Family
Inkling
Released
14 Jul 2026
Weights
Open: download and run it yourself
License
Apache-2.0
Parameters
975B (41B active)
GPU memory to run locally
~557 GB (4-bit ≈ 975 B × 0.57 GB + 1.5 GB ≈ 557 GB, estimate.)
API price
$1 in / $4.05 out per 1M tokens (OpenRouter listed price)
Access
Free (open weights)
Last verified
19 Oct 2026

Price from openrouter.ai, 19 Oct 2026.

Benchmark results

BenchmarkResultEntry on the leaderboardAs ofSource
Artificial Analysis Intelligence Index25Inkling (Xhigh)
Raw value from the page's embedded model data; the site displays it rounded to an integer.
19 Oct 2026artificialanalysis.ai
Artificial Analysis MMMU-Pro (independent evaluation)73.5%Inkling (Xhigh)
Raw fraction 0.7347 from the model data embedded in https://artificialanalysis.ai/leaderboards/models, converted to percent; rank 57 of 124 entries with an MMMU-Pro result; no last-updated date shown.
19 Oct 2026artificialanalysis.ai
AutomationBench-AA (Artificial Analysis)5%Inkling (Xhigh)
AutomationBench-AA Score (share of each task's objectives completed, a task with any guardrail violation scoring zero); raw value from the page's embedded model data (displayed rounded); rank 32 of 32; Tasks Completed 0.30%, Objectives Completed 5.84%; no last-updated date shown, so as_of = retrieval date.
19 Oct 2026artificialanalysis.ai
Design Arena SVG1,111Inkling
Design Arena SVG, rank 85, 1276 battles; rows read from POST https://www.designarena.ai/api/leaderboard (arenaType 'models', category 'svg'); names from https://www.designarena.ai/api/registry
19 Oct 2026designarena.ai
EQ-Bench Creative Writing v31,611thinkingmachines/Inkling
Elo score from the leaderboard data embedded in https://eqbench.com/creative_writing.js (file last modified 2026-09-24); rank 47 of 140; rubric score 16.38.
24 Sep 2026eqbench.com
LMArena Text Arena (overall, style control)1,441inkling
From LMArena's official leaderboard-dataset (text_style_control, split latest, category overall, published 2026-10-15), which mirrors https://arena.ai/leaderboard/text. Rank 97; 40648 votes; 95% CI 1436.6-1445.2.
15 Oct 2026huggingface.co
LMArena Text Arena: Creative Writing (style control)1,381inkling
From LMArena's official leaderboard-dataset (text_style_control, split latest, category creative_writing, published 2026-10-15), which mirrors https://arena.ai/leaderboard/text/creative-writing. Rank 144 of 414; 8748 votes; 95% CI 1373.5-1388.5.
15 Oct 2026huggingface.co
LMArena WebDev / Code Arena (overall)1,414Inkling
From LMArena's official leaderboard-dataset (webdev = Code Arena, split latest, category overall, published 2026-10-18), which mirrors https://arena.ai/leaderboard/webdev. Rank 88; 15683 votes; 95% CI 1407.4-1419.7.
18 Oct 2026huggingface.co
SWE-Bench Pro V2 (public set, full)89.9%Inkling (mini-swe-agent) xhigh
SWE-Bench Pro V2 Full (642 tasks); row added 2026-10-07; 95% CI +/-2.1; harness in the variant name; read from the leaderboard page's embedded data.
19 Oct 2026labs.scale.com

Where you can use it

Sources