Vibe check

A pelican on a bicycle

One prompt, one API key, several models. Same request body every time — only the model id changes.

Prompt, verbatim

Generate an SVG of a California brown pelican riding a bicycle. The bicycle must have spokes and a correctly shaped bicycle frame. The pelican must have its characteristic large pouch, and there should be a clear indication of feathers. The pelican must be clearly pedaling the bicycle. The image should show the full breeding plumage of the California brown pelican.

Endpoint
/v1/chat/completions
Temperature
1
Max output tokens
32,000
System prompt
none
Reasoning effort
medium / low / high
deepseek/deepseek-v4.1-flash
13,783 output tokens·medium effort
openai/gpt-6-astra
5,431 output tokens·medium effort
gemini/gemini-3.8-flash
12,047 output tokens·medium effort
anthropic/claude-fable-5-1
6,637 output tokens·medium effort
tencent/hy4-preview
23,236 output tokens·high effort
qwen/qwen3.8-flash
8,268 output tokens·medium effort
z-ai/glm-5.3-flash
3,038 output tokens·low effort
deepseek/deepseek-v4-flash-vision-exp
21,442 output tokens·medium effort
GLM-5.310.2 KB
z-ai/glm-5.3
4,638 output tokens·low effort
deepseek/deepseek-v4-pro
19,457 output tokens·medium effort
gemini/gemini-3.7-flash
14,333 output tokens·medium effort
Grok 4.612.6 KB
x-ai/grok-4.6
5,282 output tokens·medium effort
qwen/qwen3.8-max
8,888 output tokens·medium effort
deepseek/deepseek-v4-flash
29,530 output tokens·medium effort
anthropic/claude-opus-5
7,988 output tokens·medium effort
gemini/gemini-3.5-flash-lite
4,939 output tokens·medium effort
gemini/gemini-3.6-flash
11,386 output tokens·medium effort
qwen/qwen3.7-flash
10,574 output tokens
Kimi K38.3 KB
moonshot/kimi-k3
3,613 output tokens·low effort
openai/gpt-5.6-luna
3,363 output tokens·medium effort
openai/gpt-5.6-sol
6,593 output tokens·medium effort
openai/gpt-5.6-terra
5,872 output tokens·medium effort
Grok 4.56.8 KB
x-ai/grok-4.5
2,931 output tokens·medium effort
Hy34.6 KB
tencent/hy3
5,552 output tokens·low effort
anthropic/claude-sonnet-5
5,332 output tokens·medium effort
GLM-5.222.6 KB
z-ai/glm-5.2
107,412 output tokens
moonshot/kimi-k2.7-code
6,510 output tokens
anthropic/claude-fable-5
4,787 output tokens·medium effort
qwen/qwen3.7-plus
5,790 output tokens
MiniMax-M315.0 KB
minimax/minimax-m3
29,683 output tokens
anthropic/claude-opus-4-8
3,859 output tokens·medium effort
qwen/qwen3.7-max
7,088 output tokens
gemini/gemini-3.5-flash
9,930 output tokens·medium effort
qwen/qwen3.6-flash
6,950 output tokens
GPT-5.511.1 KB
openai/gpt-5.5
5,298 output tokens·medium effort
openai/gpt-5.5-pro
25,054 output tokens·medium effort
MiMo V2.57.4 KB
xiaomi/mimo-v2.5
15,873 output tokens·medium effort
xiaomi/mimo-v2.5-pro
27,559 output tokens·medium effort·64,000 max output tokens
Kimi K2.68.7 KB
moonshot/kimi-k2.6
4,439 output tokens
Grok 4.36.5 KB
x-ai/grok-4.3
2,548 output tokens·medium effort
anthropic/claude-opus-4-7
3,050 output tokens·medium effort
x-ai/grok-build-0.1
2,712 output tokens
GLM-5.123.2 KB
z-ai/glm-5.1
32,000 output tokens
qwen/qwen3.6-plus
9,213 output tokens
minimax/minimax-m2.7
4,657 output tokens
openai/gpt-5.4-mini
4,307 output tokens·medium effort
openai/gpt-5.4-nano
5,991 output tokens·medium effort
z-ai/glm-5-turbo
30,951 output tokens
GPT-5.410.3 KB
openai/gpt-5.4
7,680 output tokens·medium effort
gemini/gemini-3.1-pro-preview
4,125 output tokens·medium effort
anthropic/claude-sonnet-4-6
46,715 output tokens·medium effort·128,000 max output tokens
GLM-511.3 KB
z-ai/glm-5
24,980 output tokens
GLM-4.76.9 KB
z-ai/glm-4.7
4,460 output tokens
gemini/gemini-3-flash-preview
1,956 output tokens·medium effort
anthropic/claude-haiku-4-5
4,229 output tokens
gemini/gemini-2.5-flash-lite
3,361 output tokens·medium effort
gemini/gemini-2.5-flash
3,883 output tokens·medium effort
openai/gpt-4o-mini
921 output tokens·16,384 max output tokens

How these were made

Every drawing on this page came from the same request, sent to 58 models through one CrossModel key. Nothing else changed between them — not the endpoint, not the SDK, not the auth header.

curl https://api.crossmodel.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer cm-YOUR_KEY" \
  -d '{
  "model": "anthropic/claude-fable-5",
  "messages": [
    {
      "role": "user",
      "content": "Generate an SVG of a California brown pelican riding a bicycle. The bicycle must have spokes and a correctly shaped bicycle frame. The pelican must have its characteristic large pouch, and there should be a clear indication of feathers. The pelican must be clearly pedaling the bicycle. The image should show the full breeding plumage of the California brown pelican."
    }
  ],
  "temperature": 1,
  "max_tokens": 32000
}'

Whose test this is

The test is not ours: the prompt below is v2 of Simon Willison's pelican-riding-a-bicycle benchmark, quoted verbatim — he has run it against every new model since 2024, and revised it in November 2025 to demand a correctly built bicycle and a California brown pelican in breeding plumage. We reproduce it here through our own gateway; the parameters are ours.

Reasoning effort is set explicitly rather than left at each model’s default, which ranges from medium to xhigh — effort is the biggest lever on this task, so leaving it unset would have made the runs incomparable in the one dimension that matters most. Models with no reasoning mode get no such setting. The output budget above is the one every model started from: a model whose own ceiling is lower was clamped down to it, and two that spent the entire budget reasoning without drawing anything were given a larger one. Either way the card says which figure that model actually ran with.

This is not a benchmark

There is no score here, and nothing on this page is measured. A drawing tells you how a model handles an open-ended task on one particular day; it does not rank anything. For published, source-attributed scores, see the benchmark section on each model page.

Generated 2026-09-14. These are stored results — nothing is generated when you open this page.