Bonus: Top up now and we'll double your first deposit — get x2 credits instantly.
Get Started
Sign Up
Sign Up

chandra-ocr-2: Self-Hosting & Deployment Guide

chandra-ocr-2 is an open-weight vision-language model that reads images alongside text.

Browse available models

What is chandra-ocr-2?

chandra-ocr-2 is an open-weight vision-language model that reads images alongside text, published by datalab-to with open weights at datalab-to/chandra-ocr-2. LLM API does not route this model today, so there is no endpoint or code sample for it here. What this page does give you is the self-hosting picture: published weight sizes, the VRAM each quantisation needs and the commands to serve it yourself. Every figure is read from the public repository and refreshed weekly. At import it had 2,479,114 downloads on Hugging Face in the previous 30 days, under the openrail licence.

How to run chandra-ocr-2 yourself

Open weights (openrail), 5.3B parameters. Below: how much memory the weights need at each quantization, which hardware fits them, and the commands to serve it.

chandra-ocr-2 VRAM requirements

Precision / quantizationWeightsFigureSource
BF16 (full precision)10.6 GBEstimate5.3B params × 2 bytes
FP86.5 GBFile sizedangvansam/chandra-ocr-2-FP8-dynamic
AWQ 4-bit3.8 GBFile sizeSohailhosseini/chandra-ocr-2-AWQ-W4A16
MLX 4-bit3 GBFile size1qh/chandra-ocr-2-4bit-mlx

File size = actual download size on Hugging Face. Estimate = parameter count × bytes per weight. Weights only — the KV cache for long contexts needs extra memory; the fit tables below add 15% headroom as a rule of thumb.

GPU requirements: deploy chandra-ocr-2 on your hardware

HardwareBF16 (full precision)FP8AWQ 4-bitMLX 4-bit
NVIDIA H200141 GB1×1×1×1×
NVIDIA H10080 GB1×1×1×1×
NVIDIA A100 80GB80 GB1×1×1×1×
NVIDIA L40S48 GB1×1×1×1×

GPUs needed per variant.

HardwareMLX 4-bit
Mac, 512 GBM3 UltraFits
Mac, 192 GBM2 UltraFits
Mac, 128 GBM4 Max / M3 MaxFits
Mac, 64 GBM4 Pro / MaxFits

Assumes ~75% of unified memory is usable by the GPU.

HardwareBF16 (full precision)FP8AWQ 4-bitMLX 4-bit
AWS p5.48xlarge8× H100 · 640 GBFitsFitsFitsFits
AWS p4de.24xlarge8× A100 80GB · 640 GBFitsFitsFitsFits
AWS g6e.12xlarge4× L40S · 192 GBFitsFitsFitsFits
Google Cloud a3-highgpu-8g8× H100 · 640 GBFitsFitsFitsFits
Azure ND H100 v58× H100 · 640 GBFitsFitsFitsFits

Whole-instance GPU memory; public instance specs.

Serve it yourself

pip install -U vllm
vllm serve datalab-to/chandra-ocr-2
pip install 'sglang[all]'
python -m sglang.launch_server --model-path datalab-to/chandra-ocr-2
pip install -U mlx-lm
mlx_lm.generate --model 1qh/chandra-ocr-2-4bit-mlx --prompt "Hello"

Commands use the official repositories above; vLLM and SGLang expose an OpenAI-compatible endpoint.

Not on LLM.API yet

We do not route this model through our API at the moment. Browse the models you can call today — most workloads have a close match already live.

Browse available models

Not available on LLM API yet

We do not route this model through our API at the moment, so there is no endpoint or code snippet for it yet. Browse the models you can call today — most workloads have a close match already live.

Browse available models

When to Use — When NOT to Use

Use it if...

  • You want to run chandra-ocr-2 on your own hardware and keep every prompt inside your network
  • The weights are published openly, so you can pin one checkpoint and keep it reproducible
  • You want to fine-tune or quantise the model rather than accept a hosted configuration
  • You process documents, screenshots or charts where the image must stay in-house

Avoid if...

  • LLM API does not route this model today, so there is no endpoint here to call
  • You have no GPU capacity: the VRAM figures below are the floor, not a suggestion
  • You want zero operations — serving, scaling and upgrades are yours to run
  • You need audio or video input; this checkpoint documents text and images only

Get one key to every model

Swap your API key. Keep your code.

Browse available models Talk to sales