Skip to content
Weights Not Walls
Menu

thinkingmachines

Inkling

Reviewedaccepts images
Size
952B parameterscounted from the safetensors header on the Hub
Licence
custom, read it, Apache License 2.0Our reading, check the licence: A custom licence we have not read for this model: other makers use the same name for different texts. Read the model's own licence file before you use it.Licence text
On the Hub since
14 Jul 2026the date the repo was created
Context length
512k tokensas an API provider reports it
Kind
Base model (no base model declared on the Hub)InklingForConditionalGeneration
On Hugging Face
thinkingmachines/Inkling1.4M downloads in the last 30 days, 1.9k likes

Scores

18 independent results on 18 eval sets
Chess PuzzlesRe-run by Epoch AI
  • effort xhigh21.0%±4.113 Oct 2026, log

Epoch's name: Inkling. Best closed model on this set: claude-fable-5-2_max, 79.0% (Anthropic). All 166 results

  • effort xhigh4.9%±3.46 Aug 2026, log

Epoch's name: Inkling. Best closed model on this set: claude-fable-5-2_max, 100.0% (Anthropic). All 83 results

  • effort xhigh33.3%±2.86 Aug 2026, log

Epoch's name: Inkling. Best closed model on this set: claude-fable-5-2_max, 95.1% (Anthropic). All 127 results

GPQA diamondRe-run by Epoch AI
  • effort xhigh88.3%±1.95 Aug 2026, log

Epoch's name: Inkling. Best closed model on this set: gpt-6.2-astra_max, 96.8% (OpenAI). All 341 results

OTIS Mock AIME 2024-2025Re-run by Epoch AI
  • effort xhigh88.9%±3.95 Aug 2026, log

Epoch's name: Inkling. Best closed model on this set: claude-fable-5_high, 100.0% (Anthropic). All 318 results

SimpleQA VerifiedRe-run by Epoch AI
  • effort xhigh40.3%±1.627 Aug 2026, log

Epoch's name: Inkling. Best closed model on this set: claude-fable-5-2_max, 94.0% (Anthropic). All 105 results

APEX-AgentsReported

Epoch's name: Inkling. Best closed model on this set: gemini-4-carbon_unknown, 84.2% (Google DeepMind). All 66 results

ARC-AGIReported

Epoch's name: Inkling. Best closed model on this set: claude-fable-5_max, 98.5% (Anthropic). All 318 results

ARC-AGI-2Reported

Epoch's name: Inkling. Best closed model on this set: gpt-6-astra_max, 95.0% (OpenAI). All 299 results

CritPtReported

Epoch's name: Inkling. Best closed model on this set: gemini-4-carbon_high, 36.0% (Google DeepMind). All 271 results

DTBenchReported

Epoch's name: Inkling. Best closed model on this set: claude-fable-5-2_unknown, 99.2% (Anthropic). All 241 results

FrontierCodeReported
  • Harness: mini-swe-agent; Reasoning effort: 0.9914.0%cognition.com

Epoch's name: Inkling. Best closed model on this set: claude-opus-5-5_medium, 54.6% (Anthropic). All 46 results

FrontierSWEReported

Epoch's name: Inkling. Best closed model on this set: claude-fable-5-2_max, 69.9% (Anthropic). All 25 results

LMCAReported

Epoch's name: Inkling. Best closed model on this set: claude-fable-5-2_xhigh, 70.6% (Anthropic). All 203 results

ProofBenchReported

Epoch's name: Inkling. Best closed model on this set: claude-fable-5-1_max, 100.0% (Anthropic). All 85 results

SimpleBenchReported

Epoch's name: Inkling. Best closed model on this set: claude-fable-5-2_unknown, 94.9% (Anthropic). All 125 results

WeirdMLReported

Epoch's name: Inkling. Best closed model on this set: gpt-6-astra_promax, 93.6% (OpenAI). All 173 results

WeirdML v3Reported

Epoch's name: Inkling. Best closed model on this set: claude-fable-5-2_xhigh, 65.3% (Anthropic). All 27 results

Claimed by the maker Claimed

The model card has no claimed scores in a machine-readable form (the card's model-index), so none are shown. The maker may give scores in the card text: read the card.

Other views, which we link to and do not copy: look this model up on Artificial Analysis and the Arena leaderboard.

Independent scores from Epoch AI (CC-BY 4.0), fetched 19 Nov 2026: Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource]. Each row links its own source. What the badges mean.

Run it yourself

file sizes read from the Hub 19 Nov 2026

Memory needed = weights + context cache + 0.50 GB for the runtime. Context cache at 8k tokens: 2.1 GB, worked out from the model's config (thinkingmachines/Inkling), at 16-bit precision. Choose a machine above to see a verdict for each format. Method

GGUF for llama.cpp, Ollama, LM Studio and others

unsloth/inkling-GGUF (converted by unsloth), 70k downloads. Sizes are the real files on the Hub, all parts added up. For image input you also need the projector file mmproj-BF16.gguf (0.17 GB), not counted below.

FormatSizeMemory neededGet it
UD-IQ1_S251.6 GBfile size, 7 parts254.2 GB251.6 + 2.1 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/inkling-GGUF:UD-IQ1_S
Download
hf download unsloth/inkling-GGUF --include "UD-IQ1_S/inkling-UD-IQ1_S-*-of-00007.gguf" --local-dir inkling-GGUF
UD-IQ1_M265.5 GBfile size, 7 parts268.0 GB265.5 + 2.1 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/inkling-GGUF:UD-IQ1_M
Download
hf download unsloth/inkling-GGUF --include "UD-IQ1_M/inkling-UD-IQ1_M-*-of-00007.gguf" --local-dir inkling-GGUF
UD-Q2_K_XL295.5 GBfile size, 8 parts298.1 GB295.5 + 2.1 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/inkling-GGUF:UD-Q2_K_XL
Download
hf download unsloth/inkling-GGUF --include "UD-Q2_K_XL/inkling-UD-Q2_K_XL-*-of-00008.gguf" --local-dir inkling-GGUF
UD-Q3_K_XL403.1 GBfile size, 10 parts405.6 GB403.1 + 2.1 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/inkling-GGUF:UD-Q3_K_XL
Download
hf download unsloth/inkling-GGUF --include "UD-Q3_K_XL/inkling-UD-Q3_K_XL-*-of-00010.gguf" --local-dir inkling-GGUF
UD-Q4_K_XL546.7 GBfile size, 14 parts549.3 GB546.7 + 2.1 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/inkling-GGUF:UD-Q4_K_XL
Download
hf download unsloth/inkling-GGUF --include "UD-Q4_K_XL/inkling-UD-Q4_K_XL-*-of-00014.gguf" --local-dir inkling-GGUF
Q8_0797.9 GBfile size, 19 parts800.5 GB797.9 + 2.1 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/inkling-GGUF:Q8_0
Download
hf download unsloth/inkling-GGUF --include "Q8_0/inkling-GGUF-Q8_0-*-of-00019.gguf" --local-dir inkling-GGUF
BF161764.2 GBfile size, 41 parts1766.7 GB1764.2 + 2.1 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/inkling-GGUF:BF16
Download
hf download unsloth/inkling-GGUF --include "BF16/inkling-GGUF-BF16-*-of-00041.gguf" --local-dir inkling-GGUF

MLX for Apple silicon

FormatSizeMemory neededRepo
MLXavlp12/Inkling-975B-Alis-MLX-Dynamic-3.7bpw408.7 GBfile size411.2 GB408.7 + 2.1 + 0.50Download page
MLX 4bitpipenetwork/Inkling-MLX-4bit496.3 GBfile size498.9 GB496.3 + 2.1 + 0.50Download page

Original weights size from the safetensors header, for vLLM, SGLang, Transformers

FormatSizeMemory neededRepo
BF16original weights1773.9 GBfile size1776.5 GB1773.9 + 2.1 + 0.50
hf download thinkingmachines/Inkling --local-dir Inkling
Other conversions (9; sizes not read)

The Hub tags 48 quantised repos as conversions of this model; we read the most downloaded and show the maker's own first.

Run it by API

Hugging Face Inference Providers, fetched 19 Nov 2026, 13:58 UTC

2 providers serve this model through the Hugging Face router. Prices are US dollars per million tokens. Precision is what the provider reports; a lower precision than the original weights can change answers. Live first, then cheapest output.

ProviderInOutContextPrecisionTokens/sFirst tokenUptime 24hToolsJSON
deepinfra$0.95$4.05512knot given1300.4 s?yesyes
Together$1.00$4.05512knot given1180.2 s99.9%yesno

Through the router, with a Hugging Face token (the router bills your Hugging Face account). Change the name after the colon to pick another provider.

curl https://router.huggingface.co/v1/chat/completions \ -H "Authorization: Bearer $HF_TOKEN" -H "Content-Type: application/json" \ -d '{"model": "thinkingmachines/Inkling:deepinfra", "messages": [{"role": "user", "content": "Hello"}]}'

Not built yet: one key for every open model, with BlockRun, and our own tested-quality score per provider. Until then, precision and measured speed are the best guide to a provider here.

Planned for this page Not built yet

Try it in private chat
A link to Solana AI's confidential chat for the models it serves. This model is not listed as served there.
Fine-tune it
Fine-tuning a few base models from this page and publishing the result here.
Creators and earnings
Pages for people who publish fine-tunes and adapters of this model, and what they earn.
Fund and bounties
Bounties for new open models, paid from a public fund. How the fund will work.
Contribute compute
How to give compute to open model work, with Akash.

Model data from the Hugging Face Hub 19 Nov 2026, 13:43 UTC. The repo. Found a mistake on this page? Tell us.