Can CogVLM2 19B run on RX 6900 XT 16GB?

YES — With Offload

A82Great

Estimated from fit model

CogVLM2 19B needs ~16.5 GB VRAM. RX 6900 XT 16GB has 16.0 GB. With Q4_K_M quantization, expect ~19 tok/s.

Runtime: llama.cppCapacity: OffloadBandwidth: MediumStack: StandardBottleneck: Balanced

Operating mode

Choose the run profile you care about

Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.

Current mode

Balanced

Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.

Capabilities:

Select quantization to explore

Q4_K_M (Medium quality) — 16.5 GB, 19.0 tok/s, Runs with offload (needs ~0.4 GB host RAM)

16.5 GB required16.0 GB available

103% VRAM needed

0.5 GB over capacity — needs offload or smaller quantization

Fit status

Runs with offload (needs ~0.4 GB host RAM)

Decode

19.0 tok/s

TTFT

10215 ms

Safe context

Memory

16.5 GB / 16.0 GB

Memory breakdown

Weights11.6 GB

KV Cache2.4 GB

Runtime0.9 GB

Headroom1.6 GB

See how fast it feels

See how fast it feelsCogVLM2 19B on RX 6900 XT 16GB
1st promptCold start — includes initialization
>What is local AI inference?
Local AI inference means running an AI language model directly on your own hardware — your laptop, desktop, or server — instead of sending requests to a remote cloud API.

When you run inference locally the model weights are loaded into your GPU or unified memory. Each token you generate requires reading those weights from memory, so memory bandwidth is the main bottleneck for decode speed.

Key benefits of running locally:
  - Full privacy: your prompts never leave your machine
  - No per-token cost or rate limits
  - Works offline once the model is downloaded
  - Latency depends only on your hardware
2nd promptWarm — model ready, no init overhead
>How much VRAM do I need?
It depends on the model size and quantization level. A rough rule of thumb:

  Model size     Q4 (4-bit)    Q8 (8-bit)    FP16
  7B params      ~4.3 GB       ~7.5 GB       ~14 GB
  13B params     ~7.9 GB       ~13.9 GB      ~26 GB
  70B params     ~42.7 GB      ~74.9 GB      ~140 GB

Most people use 4-bit quantization (Q4_K_M) which gives 90-95% of full quality at a fraction of the memory. A 24 GB GPU can comfortably run most 7B-13B models.
Estimated: 19.0 tok/s decode · 10.2s TTFT (warm) · 47 tok/s prefill

What limits this setup

This setup is broadly balanced for this model.

Very little memory headroom

You can run the model, but there is not much room left for longer context, bigger batches, extra apps, or future model updates.

Best improvement path

Buy headroom, not only minimum fit

A slightly larger memory tier gives you safer context growth and makes the recommendation more future-proof.

Performance by workload

Workload	Grade	Fit	Decode	TTFT	Context
Chat	A	Runs with offload	27.1 tok/s	3901 ms	8K
Coding	A	Runs with offload (needs ~0.4 GB host RAM)	19.0 tok/s	10215 ms	8K
Agentic Coding	B	Very compromised (needs ~1.8 GB host RAM)	14.2 tok/s	19856 ms	8K
Reasoning	A	Runs with offload (needs ~0.4 GB host RAM)	19.0 tok/s	12072 ms	8K
RAG	B	Very compromised (needs ~1.8 GB host RAM)	14.2 tok/s	24820 ms	8K

Quantization options

How CogVLM2 19B (19B params) fits at each quantization level on RX 6900 XT 16GB (16.0 GB usable).

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	7.4 GB	Low	A85
Q3_K_S	3	9.3 GB	Low	A84
NVFP4	4	10.6 GB	Medium	A84
Q4_K_MBest for your GPU	4	11.6 GB	Medium	A84
Q5_K_M	5	13.7 GB	High	F0
Q6_K	6	15.6 GB	High	F0
Q8_0	8	20.3 GB	Very High	F0
F16	16	38.9 GB	Maximum	F0

Get started

Copy-paste commands to run CogVLM2 19B on your machine.

Run

docker run --rm -it ghcr.io/ggerganov/llama.cpp:full \
  --hf-repo "THUDM/cogvlm2-llama3-chat-19B" \
  --hf-file "cogvlm2-llama3-chat-19B-Q4_K_M.gguf" \
  -c 4096 -ngl 99

Your hardware

More models your RX 6900 XT 16GB can run

Model	Params	Grade	Decode	Capabilities
GPT-OSS 20B	21B	A	33.8 tok/s
Codestral 2 25.08	22B	A	12.3 tok/s

Frequently asked questions

Can RX 6900 XT 16GB run CogVLM2 19B?

Yes, RX 6900 XT 16GB can run CogVLM2 19B with a A grade (Runs with offload (needs ~0.4 GB host RAM)). Expected decode speed: 19.0 tok/s.

How much VRAM does CogVLM2 19B need?

CogVLM2 19B (19B parameters) requires approximately 16.5 GB of memory with Q4_K_M quantization.

What is the best quantization for CogVLM2 19B?

The recommended quantization for CogVLM2 19B is Q4_K_M, which balances quality and memory efficiency.

What speed will CogVLM2 19B run at on RX 6900 XT 16GB?

On RX 6900 XT 16GB, CogVLM2 19B achieves approximately 19.0 tokens per second decode speed with a time-to-first-token of 10215ms using Q4_K_M quantization.

Can RX 6900 XT 16GB run CogVLM2 19B for coding?

For coding workloads, CogVLM2 19B on RX 6900 XT 16GB receives a A grade with 19.0 tok/s and 8K context.

What context window can CogVLM2 19B use on RX 6900 XT 16GB?

On RX 6900 XT 16GB, CogVLM2 19B can safely use up to 8K tokens of context. The model's official context limit is 8K, but available memory constrains the safe maximum.

What should I upgrade first if CogVLM2 19B feels slow on RX 6900 XT 16GB?

Buy headroom, not only minimum fit. A slightly larger memory tier gives you safer context growth and makes the recommendation more future-proof.

See all results for RX 6900 XT 16GB See all hardware for CogVLM2 19B

Embed this result▼

Paste this snippet into any page to show a live fit card.

<iframe src="https://willitrunai.com/embed/cogvlm2-19b-on-rx-6900-xt-16gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>

Preview: