ZMIME
Thinking Machines · Released Nov 20, 2026

TML-d24-varloop-research

Researchers can vary the number of middle-layer passes with a single checkpoint.

  • Open weights
  • Vision
Capability
—No ECI score yet
Input
—No listed price
Output
—No listed price
Context
—tokens
Pricing

What it costs

No per-token price is published for this model yet.

No per-token price is listed for this model on models.dev.

Capability

Independent benchmarks

Epoch AI has not published a Capabilities Index score for TML-d24-varloop-research yet.

Specifications

The details

Lab
Thinking Machines
Task
Variable-depth reasoning research
Published capabilities
Researchers can vary the number of middle-layer passes with a single checkpoint.
Model size
26.4B total / 3.4B active parameters
Access
Downloadable research checkpoint
Operating limits
Native mixed-precision checkpoint occupies approximately 20 GB.
Released
Nov 20, 2026
Context window
— tokens
Inputs
Text, Images
Output
Text
Reasoning
Unconfirmed
Tool calling
Unconfirmed
Structured output
Unconfirmed

Thinking Machines model documentation

Same lab

More from Thinking Machines

Questions

About TML-d24-varloop-research

How much does TML-d24-varloop-research cost?

TML-d24-varloop-research has no published per-token API price on models.dev yet.

How good is TML-d24-varloop-research?

Epoch AI has not published a Capabilities Index score for TML-d24-varloop-research yet.

Is TML-d24-varloop-research open source?

Yes. Thinking Machines publishes the weights under the Apache-2.0 licence, so it can be downloaded and self-hosted.

What inputs does TML-d24-varloop-research support?

TML-d24-varloop-research accepts text and images. Researchers can vary the number of middle-layer passes with a single checkpoint.

When was TML-d24-varloop-research released?

Thinking Machines released TML-d24-varloop-research on Nov 20, 2026.

What are the best alternatives to TML-d24-varloop-research?

The closest current models from other labs on capability, price and release date are Hy4 Base (Tencent), Olala Mini 7A1B (OVHcloud), Shot Reader 8B (Voyager) and Aion 1.0 Instruct 2.3B (Microsoft).