Skip to content
OSMEXO
Sign inGet an API key
☰
Sign inGet an API key

Inkling

thinking-machines/inkling
  • Thinking Machines
  • Family · Inkling
  • Released 15 Jul 2026
  • Open weights
  • Text · Image · Audio → Text
  • Reasoning · adjustable
  • Tools · Vision
Not on OSMEXO yet
Request this model
Spec & list price
Context1M
Max output472K
Input / 1M$1.00
Output / 1M$4.054× input
Scores · standing
Intelligence25.0#97/489
LiveBench47.0#29/32
Coding52.1#63/196
Usage
Tokens / day53.2B▼ 53% 7d
Usage rank#45
Requests / day28.9K
Tool calls / req0.35
Per request · 30 days
Avg input18.5K1.2× median
Avg output1.2K1.7× median
Cache hit45%0.6× median
Reasoning share77%1.4× median
Tool-call failures5.1%2.7× median

Consumption

2.0B4.0B6.0B3 Oct17 Oct1 Nov
Peak 7.1B 28 SepLow 211.5M 30 OctRank today #45

News & trends

5 mentions this week ▲ 150%this model · videos, posts, articles · 30 days
Trending · 14 days
News · latest from Thinking Machines

Benchmarks

bar = share of the top score · Benchmark data: Artificial Analysis
GPQA Diamond#68 of 445
87.2%
Long-context reasoning#85 of 393
77.3%
SciCode#76 of 160
47.0%
Coding index#63 of 196
52.1
Terminal-Bench 2.1#68 of 182
55.1%
τ-Bench Banking#42 of 154
29.1%
Humanity's Last Exam#88 of 465
31.9%
Intelligence index#97 of 489
25.0
Speed · median across 2 providers, latest day
Output speed#204/33457 tok/s1.2× slower than median
Fastest provider93 tok/sbest of 2 providers
First token#2491.83 s1.8× longer than median
First answer token#4713 sthinking included

#24 cheapest of 30 with a similar score

Value
Intelligence Index
202224262830$0.5$1$2$5Inkling
Blended price per 1M tokens (log)
LiveBench suites
Browser12.0
Coding64.0
Data Analysis72.8
Design58.3
Long Agentic8.5
Reasoning78.3
Research59.9

▲ Strongest: Reasoning, Data Analysis. ▼ Weakest: Long Agentic (9).

Technical details

Reasoning
on · high
Tokenizer
Other
License
Apache 2.0
Parameters
952B
Moderation
none
Downloads · likes
2.1M · 1.8K
Reasoning effort
  • max
  • high · default
  • medium
  • low
  • minimal
  • none
Request parameters 15
  • tools
  • tool_choice
  • reasoning
  • reasoning_effort
  • include_reasoning
  • temperature
  • top_p
  • top_k
  • min_p
  • max_tokens
+5 more
  • stop
  • frequency_penalty
  • logit_bias
  • presence_penalty
  • repetition_penalty

When to use

Best for

  • Audio, image and text understanding in one model
  • Data analysis and reasoning questions
  • Very long document analysis
  • Large outputs such as long reports

Think twice for

  • Long autonomous agent runs
  • Browser-driving agents
  • Design-heavy generation
  • Structured-output pipelines

Alternatives

other vendors · similar capability

Want Inkling on OSMEXO?

Sign in to request it. Requests are reviewed; a request does not guarantee availability.

Request this model on OSMEXO