LIVE

surp/free $0.00 free│ surp/best-chat $0.012 ↓│ surp/best-coding $0.034 ↑│ usdc/base $1.00│ tps 847 ↑│ ttft 120ms ↓│ cache 22.22% hist / 0 live│ models live 221│ srp pool 2.4M ↑│ surp/free $0.00 free│ surp/best-chat $0.012 ↓│ surp/best-coding $0.034 ↑│ usdc/base $1.00│ tps 847 ↑│ ttft 120ms ↓│ cache 22.22% hist / 0 live│ models live 221│ srp pool 2.4M ↑│

verified llm performance

real output TPS, TTFT, and throughput-per-dollar — measured, not claimed.

Every number on this page comes from a real streaming request sent through surp to the Surplus Intelligence marketplace. No vendor spec sheets, no extrapolated benchmarks — only observed generation throughput. This is the data the Surplus dashboard doesn't show.

why this matters: A model listing at $0.01/M might sound cheap, but if it generates at 20 tokens/second, a 1000-token response takes 50 seconds. A model at $0.06/M generating at 100 TPS delivers the same response in 10 seconds. Throughput-per-dollar is the metric that actually determines your cost-per-task. We measure it directly.

throughput leaderboard

Models ranked by throughput-value score (p50 output TPS × tokens-per-dollar). Green ≥ 80 TPS, amber 40-79, red < 40.

#	model	p50 TPS	p95 TPS	range	p50 TTFT	price /1M	throughput/$	runs
1	gpt-5.6-luna	282.4	400000.0	42.6-400000.0	4595ms	$0.0183	15.39B	75/240
2	deepseek-v4-flash	69.0	192.9	16.1-492.6	2039ms	$0.0091	7.54B	85/245
3	glm-5.2	91.5	145.6	46.9-220.6	4412ms	$0.0171	5.37B	80/240
4	llama-3.3-70b-instruct	280.1	538.4	173.6-551.7	1327ms	$0.0588	4.76B	80/240
5	deepseek-v4-pro	53.5	69.9	43.1-83.3	2560ms	$0.0301	1.77B	80/240
6	deepseek-v4-flash-0731	88.3	128.1	19.7-326.8	2478ms	$0.1104	800.0M	86/245

what "output TPS" means

Output TPS = generated completion tokens ÷ generation seconds (time from first token to last token). This is the metric LLM buyers care about: how fast the model produces text.

This is different from request RPS (requests per second), which measures how many separate requests the gateway handles. The health board tracks RPS; this page tracks output TPS.

methodology

Real requests: each run sends a streaming POST /v1/chat/completions through surp with max_tokens=400.
TTFT: time from request submission to first output token (time-to-first-token).
Generation time: wall time − TTFT (time spent generating, excluding queue/TTFT).
Output TPS: completion_tokens / generation_seconds.
Throughput-value score: p50_output_tps × (1,000,000 / price_per_1m) — higher is better.
p50/p95: percentile across all successful runs in the window.
Token counting: prefers usage.completion_tokens from the API; falls back to len(text)/4 estimate if usage is absent.
No cherry-picking: all runs (including slow ones) are recorded. Failed runs count against the failure rate.

recent observations: gpt-5.6-luna

Last 10 raw runs — verify the numbers yourself.

TTFT	wall	price/1M	status
0ms	1103ms	$0.0000	failed
0ms	1118ms	$0.0000	failed
0ms	935ms	$0.0000	failed
0ms	1307ms	$0.0000	failed
0ms	1190ms	$0.0000	failed
0ms	1047ms	$0.0000	failed
0ms	1107ms	$0.0000	failed
0ms	859ms	$0.0000	failed
0ms	874ms	$0.0000	failed
0ms	1069ms	$0.0000	failed

try it yourself

The benchmark runner is open source. Clone the repo and run:

python3 benchmark_runner.py --model deepseek-v4-flash-0731 --runs 10 --api-key sk-sur-...

You'll get the same kind of numbers we publish here. Or just make a streaming request and time it:

curl -N -X POST https://surp.ivc.lol/v1/chat/completions \
  -H "Authorization: Bearer sk-sur-..." \
  -H "Content-Type: application/json" \
  -d '{"model":"surp/direct/deepseek-v4-flash-0731",
       "messages":[{"role":"user","content":"Write 200 words about HTTP/2"}],
       "max_tokens":400,
       "stream":true}'