← All models

Inkling

Language Available Full comparison ↗

thinking-machines/inkling · by Thinking Machines Lab · mixture-of-experts

Pricing — 1 offering(s)

Thinking Machines Lab

Input tokens (64K context)

  • $1.87 / 1M tokens (input) Current 2026-07-15 → present
    Limited-time 50% discount with no published end date: $1.87/1M prefill (input) at 64K context, down from a…

    Limited-time 50% discount with no published end date: $1.87/1M prefill (input) at 64K context, down from a $3.74/1M list rate.

Output tokens (64K context)

  • $4.68 / 1M tokens (output) Current 2026-07-15 → present

    Tinker calls this the "sample" meter. Same limited-time 50% discount as the input tier applies.

Input tokens (256K context, thinkingmachines/Inkling:peft:262144)

  • $3.74 / 1M tokens (input) Historical 2026-07-15 → 2026-10-13
    Exactly 2x the 64K variant's rate — confirmed directly against the pricing table, not assumed from the 64K…

    Exactly 2x the 64K variant's rate — confirmed directly against the pricing table, not assumed from the 64K figure. Same limited-time 50% discount applies.

  • $1.87 / 1M tokens (input) Current 2026-10-14 → present
    Limited-time 50% discount with no published end date: $1.87/1M prefill (input) at 256K context, down from a…

    Limited-time 50% discount with no published end date: $1.87/1M prefill (input) at 256K context, down from a $3.74/1M list rate. Tinker now prices 256K prefill (input) the same as 64K.

Output tokens (256K context, thinkingmachines/Inkling:peft:262144)

  • $9.36 / 1M tokens (output) Historical 2026-07-15 → 2026-10-13

    Exactly 2x the 64K variant's rate — confirmed directly, not assumed.

  • $4.68 / 1M tokens (output) Current 2026-10-14 → present
    Limited-time 50% discount with no published end date: $4.68/1M sample (output) at 256K context, down from a…

    Limited-time 50% discount with no published end date: $4.68/1M sample (output) at 256K context, down from a $9.36/1M list rate. Tinker now prices 256K sample (output) the same as 64K.

Showing the active price and any recorded history. Full pricing history is available via the paid API — see API docs.

See how Inkling fits into a cost-aware routing setup

See how →

Capability profile

How Inkling rates across core capability dimensions, with the task-level evidence behind each rating.

reasoning Strong
coding Strong

SWE-bench Verified 77.6%, SWE-bench Pro (Public) 54.3%, Terminal-Bench 2.1 (Best Harness) 63.8, GDPval-AA v2 1233 — same Hugging Face model card, confirmed 2026-08-19. Vendor-published, no independent corroboration found yet.

tool use Moderate

Mixed signal on the same model card: MCP Atlas 74.1% is solid, but τ³-Banking is only 23.7% — rated moderate rather than strong to reflect the split rather than averaging it away. Confirmed 2026-08-19.

instruction following Strong

IFBench 79.8% — same Hugging Face model card, confirmed 2026-08-19.

context window Unknown
multilingual Strong

Global-MMLU-Lite 88.7% — same Hugging Face model card, confirmed 2026-08-19.

speed Unknown
cost efficiency Unknown

Ratings are estimated — limited independent data is available for this model.

Operator guidance

Now benchmark-backed on five of eight dimensions (see capability_profile) — strong on reasoning, coding, instruction-following, and multilingual, with a mixed tool-use signal. All figures are vendor-published (Thinking Machines' own model card), not yet independently corroborated by a third-party leaderboard, and speed/cost-efficiency are still unrated. The model card positions it for general-purpose multimodal-input (text/image/audio) conversational and coding-assistant use, at a currently promotion-discounted price on Tinker, the only platform serving it today. Re-evaluate once independent corroboration or non-promotional pricing become available.

Use cases

Limitations

Citations