GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
migtissera
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
migtissera
Model Tree
Qwen/Qwen3.5-9B-Base
Input Modalities
Output Modalities
Supported Functionality
A 9B model that decisively moves the base. Across the six reported metrics, Tess-4-9B improves on Qwen3.5-9B-Base every time:
The comparison includes Qwen3.5-9B-Base, Qwythos-9B, and Tess-4-9B. Scores are accuracy on a 0–1 scale in the chart. For GPQA, the solid Tess segment is the complete 24K-cap run (62%); only samples that exhausted that cap were retried at 48K, producing the lighter extension to 73%.
Tess-4's signature is how it thinks. The reasoning/thinking traces used to train it were a best-case approximation of Fable-5, produced by Opus-4.8, GPT-5.5, and GLM-5.2 working together as a team — a multi-model teacher ensemble distilled into a single, coherent reasoning style.
The result is a model that reasons prospectively — predicting, verifying, and weighing alternatives before acting — rather than narrating after the fact.
Tess-4 uses the bundled Qwen3.5-family agentic chat template with explicit <think> … </think> reasoning blocks. The model reasons first, then produces its visible answer:
text
<|im_start|>userYour prompt here<|im_end|><|im_start|>assistant<think>… the model's private reasoning …</think> … the model's answer …<|im_end|>Apply the format automatically with tokenizer.apply_chat_template(messages, add_generation_prompt=True). Keep the repository's bundled chat template intact; substituting a generic template can materially change behavior.
Tess-4-9B is a heavy reasoner on difficult questions. For demanding math, science, or agentic tasks, allow a generous generation budget. Some GPQA samples required more than 24K output tokens and benefited from a 48K cap.
This repository — full-precision weights:
| Format | Parameters | Approximate size | Best for |
|---|---|---|---|
| BF16 safetensors | 9.65B | 19.3 GB | transformers · vLLM · SGLang |
GGUF builds → migtissera/Tess-4-9B-GGUF
| File | Format | Size | Best for |
|---|---|---|---|
| Tess-4-9B-Q4_K_M.gguf | Q4_K_M | 5.63 GB | recommended for most local users |
| Tess-4-9B-Q6_K.gguf | Q6_K | 7.36 GB | excellent quality/size balance |
| Tess-4-9B-Q8_0.gguf | Q8_0 | 9.53 GB | effectively lossless for most use cases |
| Tess-4-9B-F16.gguf | F16 | 17.92 GB | unquantized GGUF · maximum fidelity |
Run the recommended Q4_K_M build directly from Hugging Face with a recent llama.cpp:
bash
llama-cli \ -hf migtissera/Tess-4-9B-GGUF:Q4_K_M \ --jinja \ -c 65536 \ -p "Inspect this code, identify the root cause, and propose a verified fix."Or download it first:
bash
hf download migtissera/Tess-4-9B-GGUF \ Tess-4-9B-Q4_K_M.gguf \ --local-dir ./tess-4-9bSee the Tess-4-9B-GGUF model card for every quant, llama.cpp server commands, memory guidance, and prompt-format details. The current GGUF release contains text-model files only; use this full-precision repository for the inherited vision path.
Use a recent transformers release with Qwen3.5 support.
python
from transformers import AutoProcessor, AutoModelForImageTextToTextimport torch model_id = "migtissera/Tess-4-9B" processor = AutoProcessor.from_pretrained( model_id, trust_remote_code=True,)model = AutoModelForImageTextToText.from_pretrained( model_id, torch_dtype=torch.bfloat16, device_map="auto", trust_remote_code=True,) messages = [ { "role": "user", "content": "Inspect this function, identify the bug, and propose a verified fix.", }] inputs = processor.apply_chat_template( messages, tokenize=True, add_generation_prompt=True, return_dict=True, return_tensors="pt",).to(model.device) outputs = model.generate( **inputs, max_new_tokens=8192, do_sample=False,) prompt_tokens = inputs["input_ids"].shape[-1]answer = processor.decode( outputs[0][prompt_tokens:], skip_special_tokens=True,)print(answer)For multimodal input, pass image and text content through AutoProcessor using the standard Qwen3.5 message structure.
Tess-4-9B is built on Qwen/Qwen3.5-9B-Base by the Qwen team — full credit to them for an outstanding base model. Tess-4 inherits its Qwen3.5 vision-language architecture and its Apache 2.0 license.
Released under the Apache License 2.0, inherited from the base model.
bibtex
@misc{tissera2026tess49b, title = {Tess-4-9B}, author = {Migel Tissera}, year = {2026}, howpublished = {\url{https://huggingface.co/migtissera/Tess-4-9B}}, note = {Built on Qwen/Qwen3.5-9B-Base}}Tess-4-9B — part of the Tess series by Migel Tissera.
<|im_start|>userYour prompt here<|im_end|><|im_start|>assistant<think>… the model's private reasoning …</think>… the model's answer …<|im_end|>llama-cli \ -hf migtissera/Tess-4-9B-GGUF:Q4_K_M \ --jinja \ -c 65536 \ -p "Inspect this code, identify the root cause, and propose a verified fix."hf download migtissera/Tess-4-9B-GGUF \ Tess-4-9B-Q4_K_M.gguf \ --local-dir ./tess-4-9bfrom transformers import AutoProcessor, AutoModelForImageTextToTextimport torch model_id = "migtissera/Tess-4-9B" processor = AutoProcessor.from_pretrained( model_id, trust_remote_code=True,)model = AutoModelForImageTextToText.from_pretrained( model_id, torch_dtype=torch.bfloat16, device_map="auto", trust_remote_code=True,) messages = [ { "role": "user", "content": "Inspect this function, identify the bug, and propose a verified fix.", }] inputs = processor.apply_chat_template( messages, tokenize=True, add_generation_prompt=True, return_dict=True, return_tensors="pt",).to(model.device) outputs = model.generate( **inputs, max_new_tokens=8192, do_sample=False,) prompt_tokens = inputs["input_ids"].shape[-1]answer = processor.decode( outputs[0][prompt_tokens:], skip_special_tokens=True,)print(answer)@misc{tissera2026tess49b, title = {Tess-4-9B}, author = {Migel Tissera}, year = {2026}, howpublished = {\url{https://huggingface.co/migtissera/Tess-4-9B}}, note = {Built on Qwen/Qwen3.5-9B-Base}}