Route order log in sign up
1 pick your primary route

Thinking Machines: Inkling

thinkingmachines/inkling · agent · text+image+audio->text · 524k ctx · added 2026-07-17
$0.7022·4.0491 /M effective hit 43% 4.38B/d

Inkling is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 41B active parameters out of 975B total. It is designed for general-purpose reasoning, coding, agentic and tool-use systems,...

view
openrouter/thinkingmachines/inkling
Thinking Machines: Inkling · agent · text+image+audio->text · 524k ctx
4.38B/d 43% $0.7022·4.0491 – 3mo
sticker
life 429 tok/d dp age
⚡ ¢ᵉ ♻ ✓ ★
together
$1.0000·4.0500·0.1700 $0.6402·4.0500ⁱ $0.7022·4.0491ᵉ 43% – – 100% ✓ 157ms 115t/s 3mo ·
baseten/fp8
$1.0000·4.0500·0.1700 · $0.6402·4.0500ⁱ $0.7935·4.0497ᵉ · 25% · – – – · ✓ 21966ms · 52t/s · ✕1mo ·
¢ ¢ⁱ
deepinfra/fp8
$0.9500·4.0500·0.1600 · $0.6075·4.0500ⁱ $0.9098·4.0499ᵉ · 5% · – – – · ✓ 1778ms · 53t/s · ✕1mo ·
vercel/thinkingmachines/inkling
Inkling · agent · text+image+pdf->text · 256k ctx
– 43% $0.7935·4.0500 – 3mo
requesty/thinkingmachines/inkling
inkling · agent · text+image->text · 65k ctx
– $1.2214·4.6800 – 3mo
hf/thinkingmachines/Inkling
Inkling · agent · text+image->text · 524k ctx
– $0.9500·4.0500 – 3mo
compare all models →