Qwen2.5-Coder-14B-Instruct
18 workload results across 4 hardware configurations.
Fastest config measured by llm-speed (@llmspeed)
137.3 decode tok/s
on RTX 5090 (32GB) via llama.cpp (Q4_K_M). see full run
For a standard Q4_K_M quantization using llama.cpp, here are measured results for short-chat generation.
On llama.cpp Q4_K_M, Qwen2.5-Coder-14B-Instruct is fastest on the RTX 5090 at 137.3 tok/s, 53% faster than the RTX 4090 (89.87 tok/s); slowest: RTX 3090 at 78.21 tok/s.
| Card | Decode tok/s | TTFT (131-token prompt) | Run | Source |
|---|---|---|---|---|
| RTX 5090 | 137.3 (single run) | 65 ms | r_xnr_k4emoyr | llm-speed (@llmspeed) |
| RTX 4090 | 89.87 (single run) | 57 ms | r_eqnsqmgz29a | community-submitted |
| RTX 3090 | 78.21 (single run) | 77 ms | r_rkyf9wxzw9z | llm-speed (@llmspeed) |
Qwen2.5-Coder-14B-Instruct, Q4_K_M, llama.cpp, chat-short, batch 1. Only cards with their standard memory are listed; CPU-only hosts are excluded. Runs use different llama.cpp versions (RTX 5090: 0.7.0-dev (build 1, commit 1268eda); RTX 4090: 0.6.0-dev (build 1, commit df18bee); RTX 3090: 0.7.0-dev (build 1, commit ea0c4c2)), so treat this as measured reference points, not a controlled ranking. Community-submitted runs are signed on the submitter's own machine and are not measured by llm-speed; llm-speed verdicts use only its own runs.
Signed runs (18 workload results)
2 of 6 runs were measured by llm-speed (@llmspeed); the other 4 are community-submitted (signed on the submitter's own machine, not measured by llm-speed, and not used for llm-speed verdicts).
RTX 5090 (32GB)
| Workload | Backend | Quant | decode tok/s | prefill tok/s | TTFT | Run | Source |
|---|---|---|---|---|---|---|---|
| chat-short | llama.cpp@0.7.0-dev (build 1, commit 1268eda) | Q4_K_M | 137.3tok/s | 2,016.9tok/s | 65.0ms | r_xnr_k4emoyr | Measured by llm-speed@llmspeed |
| chat-long | llama.cpp@0.7.0-dev (build 1, commit 1268eda) | Q4_K_M | 128.9tok/s | 5,987.2tok/s | 529ms | r_xnr_k4emoyr | Measured by llm-speed@llmspeed |
| concurrent-decode | llama.cpp@0.7.0-dev (build 1, commit 1268eda) | Q4_K_M | 136.7tok/s | no data | no data | r_xnr_k4emoyr | Measured by llm-speed@llmspeed |
| agent-trace | llama.cpp@0.7.0-dev (build 1, commit 1268eda) | Q4_K_M | 128.2tok/s | 18,881.9tok/s | 113ms | r_xnr_k4emoyr | Measured by llm-speed@llmspeed |
| chat-short | llama.cpp | - | 140.0tok/s | no data | 42.1ms | r_opuj21f13-_ | Community-submitted |
| chat-long | llama.cpp | - | 123.0tok/s | no data | 457ms | r_opuj21f13-_ | Community-submitted |
| concurrent-decode | llama.cpp | - | 133.8tok/s | no data | no data | r_opuj21f13-_ | Community-submitted |
| agent-trace | llama.cpp | - | 130.5tok/s | 21,985.9tok/s | 104ms | r_opuj21f13-_ | Community-submitted |
| chat-short | llama.cpp | - | 136.0tok/s | no data | 51.3ms | r_p_f63tcgans | Community-submitted |
RTX 3090 (24GB)measured by llm-speed
| Workload | Backend | Quant | decode tok/s | prefill tok/s | TTFT | Run | Source |
|---|---|---|---|---|---|---|---|
| chat-short | llama.cpp@0.7.0-dev (build 1, commit ea0c4c2) | Q4_K_M | 78.21tok/s | 1,693.2tok/s | 77.4ms | r_rkyf9wxzw9z | Measured by llm-speed@llmspeed |
| chat-long | llama.cpp@0.7.0-dev (build 1, commit ea0c4c2) | Q4_K_M | 74.31tok/s | 2,549.2tok/s | 1,243ms | r_rkyf9wxzw9z | Measured by llm-speed@llmspeed |
| concurrent-decode | llama.cpp@0.7.0-dev (build 1, commit ea0c4c2) | Q4_K_M | 77.11tok/s | no data | no data | r_rkyf9wxzw9z | Measured by llm-speed@llmspeed |
| agent-trace | llama.cpp@0.7.0-dev (build 1, commit ea0c4c2) | Q4_K_M | 75.86tok/s | 7,427.0tok/s | 273ms | r_rkyf9wxzw9z | Measured by llm-speed@llmspeed |
RTX 4090 (24GB)
| Workload | Backend | Quant | decode tok/s | prefill tok/s | TTFT | Run | Source |
|---|---|---|---|---|---|---|---|
| chat-short | llama.cpp@0.6.0-dev (build 1, commit df18bee) | Q4_K_M | 89.87tok/s | 2,292.6tok/s | 57.1ms | r_eqnsqmgz29a | Community-submitted |
| chat-long | llama.cpp@0.6.0-dev (build 1, commit df18bee) | Q4_K_M | 84.17tok/s | 4,989.1tok/s | 635ms | r_eqnsqmgz29a | Community-submitted |
| concurrent-decode | llama.cpp@0.6.0-dev (build 1, commit df18bee) | Q4_K_M | 88.90tok/s | no data | no data | r_eqnsqmgz29a | Community-submitted |
| agent-trace | llama.cpp@0.6.0-dev (build 1, commit df18bee) | Q4_K_M | 86.64tok/s | 14,684.0tok/s | 142ms | r_eqnsqmgz29a | Community-submitted |
M3 Ultra (60-core GPU) + 96GB unified
| Workload | Backend | Quant | decode tok/s | prefill tok/s | TTFT | Run | Source |
|---|---|---|---|---|---|---|---|
| chat-short | mlx@0.31.3 | 4BIT | 70.51tok/s | 306.1tok/s | 428ms | r_l36cijqxq4t | Community-submitted |
Can Qwen2.5-Coder-14B-Instruct run on your hardware?
Fastest decode speed on each GPU or Mac, from llm-speed's own runs and community-submitted runs. Each page lists every quant and runtime and labels which runs llm-speed measured.
Qwen2.5-Coder-14B-Instruct on hardware
Request a Qwen2.5-Coder-14B-Instruct run
No measured run yet for these pairs. Vote for the one you want; the most-requested runs go first. See all requests.
We never sell or share your email. Privacy
Compare Qwen2.5-Coder-14B-Instruct head-to-head
GPUs compared on Qwen2.5-Coder-14B-Instruct
Check it for your setup
- Predict your exact setup: estimate whether Qwen2.5-Coder-14B-Instruct fits your GPU or Mac at your quant and context, next to measured speeds where we have them.
- Request this pair: ask for a measured Qwen2.5-Coder-14B-Instruct run on hardware we have not measured. Requests are ranked by votes.
- Full run table (6 runs): every Qwen2.5-Coder-14B-Instruct result we have, with links to each source run.