Inkling
$4.21
estimated usage cost
1,000 requests × (1,000 uncached input × $1.87 + 500 output × $4.68) / 1,000,000 = $4.21
base rates; no context pricing tiers are published.
Text generation / Thinking Machines
Inkling
Inkling is a text generation model from Thinking Machines, available through the Impossibl API. Its published context window is 65,536 tokens. The advertised input rate is $1.87 per million input tokens. The advertised output rate is $4.68 per million output tokens.
api id: thinkingmachines/inkling
how much does inkling cost?
published base api rates · usd per 1m tokens
- input
- $1.87
- output
- $4.68
- cache read
- $0.374
rates are published by the impossibl api. balance-purchase fees, taxes, tool charges, and workspace-specific pricing are separate.
specifications & capabilities
- creator
- Thinking Machines
- model type
- Text generation
- context window
- 65.54k tokens
- maximum output
- not published in the catalog
- input types
- text, image
- output types
- text
- tool calling
- not verified in the catalog
unknown means the public catalog does not establish that fact. a model's documented capabilities can differ from what a particular api supports.
estimate your usage
estimate your api cost
estimate text token usage for your workload. all amounts are in usd.
estimates use published impossibl api usage rates. configured billing adjustments, funding fees, taxes, and custom workspace rates are excluded and may change the final amount.
image and audio usage, cache creation, and additional tool charges are excluded. reasoning tokens count as billable output, even when they are not visible in the response.
compare more Text generation models
- Inkling 256K$3.74 input / 1m tokenscompare →
- Claude Fable 5$10.00 input / 1m tokenscompare →
- Claude Fable 5.1$10.00 input / 1m tokenscompare →
- Claude Haiku 4.5$1.00 input / 1m tokenscompare →
independent benchmarks
Results from Artificial Analysis. Each configuration is listed separately. These evaluations describe the tested model configuration; performance and costs can differ on Impossibl. Last checked 2026-09-11.
Inkling (xhigh)
| Benchmark | Score |
|---|---|
| Intelligence Index | 25.54 |
| Omniscience Index | 2.00 |
| GPQA | 87.17% |
| Humanity’s Last Exam | 31.88% |
| SciCode | 46.99% |
| CritPt | 5.43% |
| Terminal-Bench v2.1 | 55.06% |
| Terminal-Bench v4.0 | 1.01% |
| AA-LCR | 77.33% |
| Multilingual LCR | 12.22% |
| GDPval Elo | 1165.19 |
| GDP.pdf All-pass | 12.80% |
| AutomationBench Partial | 4.98% |
| AnalystAgent | 23.75% |
| MMMU-Pro | 73.47% |
| Tau Banking | 29.07% |
| coding index (API) | 52.1 |
| agentic index (API) | 24.3 |
| intelligence index (API) | 25.5 |
Unmeasured scores are omitted. API values retain their source units.
view source ↗frequently asked questions
What is Inkling?
Inkling is a text-generation model from Thinking Machines, available through the Impossibl API.
How much does Inkling cost?
Inkling costs $1.87 per million input tokens and $4.68 per million output tokens through the Impossibl API. Cache reads cost $0.374 per million tokens. All rates are in USD.
What is the context length of Inkling?
Inkling has a 65,536-token context window through the Impossibl API.
What inputs and outputs does Inkling support?
Inkling accepts text and image as input. Inkling returns text.
How do I use Inkling through an API?
Inkling is available through the Impossibl API. Use thinkingmachines/inkling as the model identifier with /v1/chat/completions. The quickstart explains how to create an API key and send your first request.
When was Inkling released?
Thinking Machines released Inkling on July 15, 2026.
sources & coverage
this page describes Inkling as listed by the impossibl public api. prices and availability come from that catalog, refreshed here every five minutes. this is an impossibl offering, not a comparison of independent hosting-provider quotes.
model descriptions and release dates, where available, use the creator sources linked in the faq. api prices and limits use the impossibl catalog. benchmark results, where available, are credited to Artificial Analysis in the benchmarks section. report a correction →