Moving a basic image call from GPT Image 2.5 to 2.6 is a model-ID change. The endpoints are the same, the token rates are the same, and the three recommended sizes are the same. The parts that need attention are Batch, which 2.5 has and 2.6 does not; the Responses image tool, which OpenAI's 2.5 and 2.6 materials describe differently; and the controls — complexity, SVG, GIF, and markup — which are documented for OpenAI's Image API but are not in AIHubMix's published parameter schema yet.
This tutorial walks one representative job through AIHubMix: a 1536x1024 poster with a headline and a product shot. It starts with three cheap Flare drafts, picks one, renders the final with Sunburst, and puts a cost line on every step. Along the way it covers the migration problems that fail quietly rather than loudly.
What stays the same
The two 2.6 models are gpt-image-2.6-sunburst and gpt-image-2.6-flare. Both are called through v1/images/generations and v1/images/edits; neither has a plain gpt-image-2.6 alias. Text input is $5 per million tokens, image input $8, and image output $30, exactly as on 2.5 and on GPT Image 2.
That means a pipeline built around the Image API keeps its request shape. What changes is the model string, and only if you want the new model. The speed claim — 50% faster for the family, according to AIHubMix's model page — is the main reason to change on day one.
Migration gotchas
Batch is gone
This is the one feature where 2.5's advantage is unambiguous. GPT Image 2.5's model page lists v1/batch as supported, and Batch prices image output at $15 per million tokens instead of $30. GPT Image 2.6's model page lists Batch as not supported, and the Batch pricing table omits both 2.6 models.
If you run overnight image jobs, do not move them to 2.6. Either stay on 2.5 or split the workload: interactive drafts on 2.6, batch renders on 2.5. AIHubMix's image docs list no Batch endpoint, so this decision only applies to direct OpenAI calls.
Responses pipelines need a check
OpenAI's own materials disagree about 2.5 here. The September changelog lists the Responses API image generation tool alongside the Image API, but the 2.5 model page marks Responses as not supported. For 2.6 the picture is clearer: the model page marks Responses as not supported, and the November changelog lists only v1/images/generations and v1/images/edits.
If your code reaches 2.5 through Responses, add a smoke test before changing the model ID. A failure here is loud, but only if it is tested; a pipeline that silently falls back to another model is worse than a 400.
The new controls are Image API only
The image generation guide puts complexity, markup-guided edits, SVG output, and animated GIF output on the Image API. None of them is available through the Responses image tool, and SVG and GIF are generation-only: neither works on /v1/images/edits.
Each has its own constraint. SVG requires n=1 and does not support streaming, partial images, output_compression, or prompt_rewriter. GIF uses frame_count (default 9) and frame_duration_ms (default 200), is generation-only, and does not support streaming, partial images, custom output_compression, or prompt_rewriter either. Markup edits need the original image as the first input image and a separate transparent PNG of the same dimensions; neither may exceed 16,777,216 pixels, and the markup file is capped at 50 MB. Do not combine markup with mask or prompt_rewriter.
There is also a billing rule for SVG: you pay the image model's rate for the reference image and the authoring model's rate for the tokens used to produce the SVG. The total depends on both, so the usual per-image estimate does not cover it.
Quality and size documents disagree
For 2.5, OpenAI documents five explicit quality levels besides auto: low, medium, high, xhigh, and max. For 2.6, the guide's cost calculator lists the same five for both generations, the API reference attaches xhigh and max to 2.5 snapshots only, and AIHubMix's per-image table — the source of the costs below — prices just low, medium, high, and auto.
The safe set for 2.6 today is low, medium, high, and auto. The same caution applies to sizes: custom WIDTHxHEIGHT values are documented for GPT Image 2 and 2.5, while 2.6's guide recommends 1024x1024, 1536x1024, and 1024x1536. Use the three recommended sizes until the 2.6 documentation is consistent.
Coming from 1.5, 1024x1024 high quality costs more
GPT Image 1.5 and 1-mini retire on December 1, 2026, with 2.5 named as the replacement. If you render mostly at 1024x1024 high quality, the move is not cheap: OpenAI's 1.5 table puts high quality at $0.133, and AIHubMix's 2.6 estimate is $0.211, about 59% more. Medium goes from $0.034 to $0.053, and low goes the other way, from $0.009 to $0.006. At 1536x1024, 1.5 high is $0.200, so 2.6's $0.165 is cheaper at that size.
Run the same mix on 2.5 before assuming 2.6: the token rates are identical, so the difference between the two 2.x models is token consumption, not list price.
Calling 2.6 through AIHubMix
AIHubMix serves both 2.6 models and prices them at the official rates. AIHubMix's image generation docs describe the newer endpoint, POST https://aihubmix.com/ai/v1/images/generations. It accepts model, prompt, n, size, output_format, image, images, mask, async, extra, and the two webhook fields.
The important detail is where quality lives. On /ai/v1/images/generations there is no top-level quality field; it goes inside extra:
curl -s https://aihubmix.com/ai/v1/images/generations \
-H "Authorization: Bearer $AIHUBMIX_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-image-2.6-flare",
"prompt": "A minimal poster: a single ceramic coffee mug on a pale sand background, soft morning light, large empty space at the top for a headline",
"size": "1024x1024",
"n": 3,
"extra": { "quality": "low" }
}'
The response carries one or more output URLs. They are short-lived — roughly 30 minutes — and must be downloaded with the same Authorization: Bearer $AIHUBMIX_API_KEY header:
curl -s -H "Authorization: Bearer $AIHUBMIX_API_KEY" "<output-url>" -o draft-1.png
AIHubMix's docs recommend /ai/v1 for new integrations: the OpenAI-compatible /v1 endpoints may not yet support the latest models, and the legacy image reference's model list includes gpt-image-2, gpt-image-1.5, gpt-image-1-mini, and the DALL·E models, but not 2.6. The per-model schema does list a /v1/images/generations entry for 2.6, but /ai/v1 is the path AIHubMix documents for 2.6 and the one that adds async tasks and webhooks.
What the schema does not expose
AIHubMix's published schema for 2.6 accepts output_format values of png, jpeg, and webp only. It has no complexity, markup, frame_count, or frame_duration_ms, and no svg or gif output. The schema is available without authentication at https://aihubmix.com/call/schema/models/gpt-image-2.6-flare/endpoints, and it is the thing to check before relying on any of those features through the gateway.
This is a schema gap, not a measured failure: the features are documented on OpenAI's side, and the gateway may add them later. Until the schema lists them, plan SVG, GIF, and markup work as direct-to-OpenAI calls.
Errors to expect
| Status | Meaning | What to do |
|---|---|---|
| 400 | Parameter error | Check prompt, size, n |
| 401 | Missing or invalid key | Check Authorization header |
| 402 | Insufficient quota | Add credits |
| 403 | Permission denied | Check account, key, IP rules |
| 429 | Rate limited | Back off and retry |
| 503 | No channel available | Check model ID, retry later |
A 400 often comes from the provider rather than the gateway. AIHubMix's error bodies carry a tid trace id. Keep it for support, and do not retry a 400 unchanged. This is the /ai/v1 contract; the legacy /v1 endpoints report insufficient quota as 403 insufficient_user_quota instead.
The workflow: three drafts, one final
The job is a poster with a product shot and a headline. The prompt stays fixed; only the model, size, and quality change.
Step 1 — draft with Flare. One request with n=3 at 1024x1024 and low quality returns three directions. At an estimated $0.006 per image, the step costs $0.018.
Step 2 — pick one. This is the cheap part of the workflow and the part that benefits most from model speed. Three drafts in one call is a normal budget, not a splurge.
Step 3 — render the final with Sunburst. The chosen composition is re-prompted at 1536x1024 and high quality. At an estimated $0.165 per image, this is the expensive step; one final costs about 28 times one draft image, or about nine times the three-draft step.
curl -s https://aihubmix.com/ai/v1/images/generations \
-H "Authorization: Bearer $AIHUBMIX_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-image-2.6-sunburst",
"prompt": "A minimal poster: a single ceramic coffee mug on a pale sand background, soft morning light, large empty space at the top for a headline. Refined studio product photography, subtle shadow, no text",
"size": "1536x1024",
"n": 1,
"extra": { "quality": "high" }
}'
Step 4 — total. An estimated $0.018 of drafts plus $0.165 of final is $0.183 for four output images. If the final is rendered at medium quality instead, it is $0.041 and the run is $0.059. Input text tokens add a fraction of a cent at these prompt lengths; an edit would add image input tokens at $8 per million.
That number shows there is no cost difference between the two models: the same four images on 2.5 cost the same, because AIHubMix's per-image estimates match. The reason to use 2.6 is the faster draft loop, and the option to move SVG or GIF work to OpenAI later, not a lower bill.
Migration checklist
- Replace
gpt-image-2.5-sunburstandgpt-image-2.5-flarewith the 2.6 IDs where you want the new model; leave Batch jobs on 2.5. - Confirm whether any pipeline uses the Responses image tool; 2.6 documents the Image API only.
- Keep quality to
low,medium,high, orautoand sizes to the three recommended values until the 2.6 docs settle. - On AIHubMix's
/ai/v1, put quality inextra; for 2.6, use/ai/v1, the path AIHubMix recommends for new integrations. - Download output URLs within 30 minutes, with the Bearer header.
- Do not assume
complexity, markup, SVG, or GIF work through AIHubMix until the schema lists them. - If you are on 1.5, budget for the December 1 retirement and re-check the high-quality cost.
FAQ
Do I need to change my endpoint for GPT Image 2.6?
On OpenAI's API, no. It uses v1/images/generations and v1/images/edits, the same as 2.5; only the model ID changes. On AIHubMix, use /ai/v1/images/generations, as described above.
Is GPT Image 2.6 cheaper than 2.5?
No. The token rates are identical: $5 per million text input tokens, $8 per million image input tokens, and $30 per million image output tokens.
Why can't I use Batch with GPT Image 2.6?
OpenAI's 2.6 model pages mark Batch as not supported, and the Batch pricing table lists 2.5 but not 2.6. Batch is the clearest workflow where 2.5 remains the better choice.
How do I set quality on AIHubMix?
For 2.6, on /ai/v1/images/generations, put it in extra.quality. The OpenAI-compatible /v1/images/generations has a top-level quality field, but AIHubMix recommends /ai/v1 for new integrations; use /ai/v1 for 2.6.
Can I generate SVG or GIF through AIHubMix?
Not according to the current schema, which lists png, jpeg, and webp and does not include complexity, markup, frame count, or SVG/GIF output. Those are documented for OpenAI's Image API; check the schema before relying on them through the gateway.
How long do AIHubMix output URLs last?
Roughly 30 minutes. Download them with the same Bearer token and store the file yourself.
What happens to GPT Image 1.5?
It retires on December 1, 2026. OpenAI names 2.5 as the replacement, but at 1024x1024 high quality 2.6 costs more per image than 1.5 did; at 1536x1024 it costs less.
Keep reading: the GPT Image 2.6 series
The comparison post covers the price tables, the speed claim, and where 2.5 still wins; it is the place to start if you have not decided whether to move.
GPT Image 2.6 vs 2.5: is the upgrade worth it?
