skills/generate-image/references/models.md
Snapshot of GET https://openrouter.ai/api/v1/images/models, taken 2026-07-26 (38 models).
The catalogue moves; re-read it rather than trusting this table:
python scripts/generate_image.py --list-models
No API key is needed to list models.
refs is the maximum number of input_references accepted; n is images per request.
| Model | n | refs | stream | Parameters |
|---|---|---|---|---|
google/gemini-3.1-flash-image | 1 | 14 | no | aspect_ratio, resolution, n, input_references |
google/gemini-3.1-flash-image-preview | 1 | 14 | no | aspect_ratio, resolution, n, input_references |
google/gemini-3.1-flash-lite-image | 1 | 14 | no | aspect_ratio, resolution, n, input_references |
google/gemini-3-pro-image | 1 | 14 | no | aspect_ratio, resolution, n, input_references |
google/gemini-3-pro-image-preview | 1 | 14 | no | aspect_ratio, resolution, n, input_references |
google/gemini-2.5-flash-image | 1 | 3 | no | aspect_ratio, n, input_references |
black-forest-labs/flux.2-pro | 1 | 8 | no | aspect_ratio, output_format, seed, n, input_references |
black-forest-labs/flux.2-max | 1 | 8 | no | aspect_ratio, output_format, seed, n, input_references |
black-forest-labs/flux.2-flex | 1 | 8 | no | aspect_ratio, output_format, seed, n, input_references |
black-forest-labs/flux.2-klein-4b | 1 | 4 | no | aspect_ratio, output_format, seed, n, input_references |
bytedance-seed/seedream-4.5 | 10 | 14 | no | aspect_ratio, resolution, seed, n, input_references |
openai/gpt-image-2 | 10 | 16 | yes | aspect_ratio, background, quality, output_compression, n, input_references |
openai/gpt-image-1 | 10 | 16 | yes | aspect_ratio, background, quality, output_compression, n, input_references |
openai/gpt-image-1-mini | 10 | 16 | yes | aspect_ratio, background, quality, output_compression, n, input_references |
openai/gpt-5.4-image-2 | 10 | 16 | yes | background, quality, output_compression, n, input_references |
openai/gpt-5-image | 10 | 16 | yes | background, quality, output_compression, n, input_references |
openai/gpt-5-image-mini | 10 | 16 | yes | background, quality, output_compression, n, input_references |
recraft/recraft-v4.1-pro | 6 | 1 | no | aspect_ratio, n, input_references |
recraft/recraft-v4.1 | 6 | 1 | no | aspect_ratio, n, input_references |
recraft/recraft-v4.1-vector | 6 | 1 | no | aspect_ratio, n, input_references |
recraft/recraft-v4.1-pro-vector | 6 | 1 | no | aspect_ratio, n, input_references |
recraft/recraft-v4.1-utility | 6 | 1 | no | aspect_ratio, n, input_references |
recraft/recraft-v4.1-utility-pro | 6 | 1 | no | aspect_ratio, n, input_references |
recraft/recraft-v4-pro | 6 | 1 | no | aspect_ratio, n, input_references |
recraft/recraft-v4 | 6 | 1 | no | aspect_ratio, n, input_references |
recraft/recraft-v4-vector | 6 | 1 | no | aspect_ratio, n, input_references |
recraft/recraft-v4-pro-vector | 6 | 1 | no | aspect_ratio, n, input_references |
recraft/recraft-v3 | 6 | 1 | no | aspect_ratio, n, input_references |
sourceful/riverflow-v2.5-pro | 1 | 10 | no | aspect_ratio, background, output_format, resolution, n, input_references |
sourceful/riverflow-v2.5-fast | 1 | 4 | no | aspect_ratio, background, output_format, resolution, n, input_references |
sourceful/riverflow-v2-pro | 1 | 10 | no | aspect_ratio, resolution, n, input_references |
sourceful/riverflow-v2-fast | 1 | 4 | no | aspect_ratio, resolution, n, input_references |
microsoft/mai-image-2.5-pro | 1 | 1 | no | aspect_ratio, n, input_references |
microsoft/mai-image-2.5 | 1 | 1 | no | aspect_ratio, n, input_references |
krea/krea-2-large | — | 1 | no | aspect_ratio, resolution, seed, input_references |
krea/krea-2-medium | — | 1 | no | aspect_ratio, resolution, seed, input_references |
krea/krea-2-medium-turbo | — | 1 | no | aspect_ratio, resolution, seed, input_references |
x-ai/grok-imagine-image-quality | 1 | 3 | no | aspect_ratio, resolution, n, input_references |
Parameters are not universal, and an unsupported one is rejected, not ignored:
resolution (512, 1K, 2K, 4K) — Gemini, Seedream, Riverflow, Krea, Grok. Not FLUX.output_format — FLUX (png, jpeg) and Riverflow 2.5. Gemini does not take it.quality, background, output_compression — the OpenAI family (background also on Riverflow 2.5).seed — FLUX, Seedream, Krea. Not Gemini, not OpenAI.aspect_ratio — nearly everything, but the enum differs. FLUX 2 Pro accepts
1:1, 4:3, 3:4, 3:2, 2:3, 16:9, 9:16, 21:9, auto.Per-model detail, including provider-specific passthrough parameters and pricing:
curl -s https://openrouter.ai/api/v1/images/models/black-forest-labs/flux.2-pro/endpoints
FLUX 2 Pro, for example, allows steps, guidance, and safety_tolerance as passthrough
parameters and bills $0.03 per output megapixel.
google/gemini-3.1-flash-image (skill default), or
google/gemini-3-pro-image for the higher tier.black-forest-labs/flux.2-pro.black-forest-labs/flux.2-klein-4b, google/gemini-3.1-flash-lite-image.bytedance-seed/seedream-4.5 or the OpenAI family (up to 10 per request).recraft/recraft-v4-vector, recraft/recraft-v4.1-vector, and
the -pro-vector variants. These return media_type: image/svg+xml.--background transparent.The Image API accepts the same provider block as chat completions — provider.only,
provider.order, provider.ignore, provider.sort (price, throughput, latency), and
provider.allow_fallbacks. The bundled script does not expose these; send a direct request when
routing control matters.
Image billing is all-or-nothing: a generation either completes and is billed in full, or fails and
is not billed. Partial preview frames delivered during streaming are not charged separately.
Per-request cost comes back in usage.cost, which the script prints.