Tholos-2B GGUF: local AI agents with Ollama and llama.cpp

This project is by Eschatia Labs, an independent non-profit open-source AI research lab.

GGUF builds of Tholos-2B, a small language model for AI agents that choose tools for tables, notes and tasks. Tholos-2B passes 137/160 scenarios; MiniCPM5-2B passes 112/160 on Tholos-Bench with Q4_K_M, llama.cpp JSON schema decoding and one Kaggle T4. Only Tholos-2B was fine-tuned for this task format.

Quick facts

Item Value
Quantized from mertkayacs/Tholos-2B
Q4_K_M file size 1.56 GB
Q8_0 file size 2.68 GB
CPU passes (llama.cpp JSON schema) 134/160 (MiniCPM5-2B: 117/160)
CPU passes (Ollama JSON mode) 136/160 (MiniCPM5-2B: 96/160)
Injection scenarios passed (Kaggle T4) 10/12 (MiniCPM5-2B: 8/12)
Hugging Face downloads 374 (as of 2026-10-15)
GitHub stars 1

Project links

Cite this

@techreport{kaya2026tholosreport,
  author      = {Kaya, Mert},
  title       = {Tholos-2B: Technical report},
  institution = {Eschatia Labs},
  year        = {2026},
  url         = {https://eschatialabs.com/research/tholos-2b-report/},
  note        = {Technical report; peer review pending}
}

Run with Ollama

ollama pull hf.co/mertkayacs/Tholos-2B-GGUF:Q4_K_M && uv tool install git+https://github.com/mertkayacs/tholos && tholos

The same model is also in the Ollama library as mertkayacs/tholos-2b (ollama pull mertkayacs/tholos-2b).

Open http://127.0.0.1:7070, select Detect in Settings, then add the model. The repo's template and params preserve the trained prompt format, stop tokens and 16,384-token context.

Direct Ollama clients should request json_object mode and reasoning_effort: "none", then validate each step. Tholos does this for you and applies approval rules.

Run with llama.cpp

llama-server -hf mertkayacs/Tholos-2B-GGUF:Q4_K_M --jinja -c 16384 -t 4 -a tholos-2b --host 127.0.0.1 --port 8080

Use json_schema response format. The complete request shows the tool definitions, step schema and response loop.

Files

Quantization File size
Q4_K_M 1.56 GB
Q8_0 2.68 GB

Use Q4_K_M to match the benchmark. SHA256SUMS records the checksums. File size is not total runtime RAM.

Results and limits

With Q4_K_M on a CPU, Tholos-2B passes 134/160 with llama.cpp JSON schema, against 117/160 for MiniCPM5-2B. With Ollama JSON mode and the repo template, the counts are 136/160 and 96/160. Compare results within one setup.

English only. The model learned Tholos's thirteen tools; other formats are unmeasured. Validate calls, treat external text as untrusted, and keep approvals enabled for consequential actions.

Apache-2.0. Full-precision weights, training and evaluation | Code | Kaggle | Project | Benchmark page.

An Eschatia Labs project. Mert Kaya.

Downloads last month
435
GGUF
Model size
3B params
Architecture
llama
Hardware compatibility
Log In to add your hardware

4-bit

8-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Model tree for mertkayacs/Tholos-2B-GGUF

Quantized
(2)
this model

Dataset used to train mertkayacs/Tholos-2B-GGUF

Space using mertkayacs/Tholos-2B-GGUF 1

Collections including mertkayacs/Tholos-2B-GGUF