HiAPI
Blog

Guide

Ideogram 4.5 API Pricing: From $0.043 per Image at LOW

Three generation tiers, seven editor combinations, a same-prompt tier comparison, and budgets.

Oct 26, 2026 5
Ideogram 4.5 API Pricing: From $0.043 per Image at LOW
LOW / image
$0.043
MEDIUM default
$0.086
HIGH / image
$0.3145

TL;DR

  • The Ideogram 4.5 family (ideogram-v4.5/text-to-image for generation, ideogram-v4.5/image-to-image for editing) bills per image by quality tier: LOW $0.043, MEDIUM (default) $0.086, HIGH $0.3145; the editor adds VERY_LOW at $0.0115. See the full rate table.
  • Each request currently returns 1 image; num_images must be 1. Four separate LOW tasks total $0.172; omitting quality bills at MEDIUM.
  • Both are live: the generation page and the Edit page. Usage is in the API guide.

Price table: generation, three tiers

Tier (quality) Per image 100 images 1,000 images $10 yields
LOW $0.043 $4.30 $43.00 232
MEDIUM (default) $0.086 $8.60 $86.00 116
HIGH $0.3145 $31.45 $314.50 31

HiAPI current prices, checked October 26, 2026.

Generation size and prompt expansion do not change the selected quality-tier rate. References and masks on the editing endpoint do not add a fee either. Each task returns 1 image: four separate LOW tasks total $0.172, four MEDIUM tasks $0.344, and two HIGH tasks $0.629. A validation rejection before task creation does not start generation; for a created task, check its final status and the account record when reconciling the charge. Omitting quality bills at MEDIUM — pass low explicitly to save.

Same prompt, three tiers, real outputs

The following outputs illustrate three quality tiers with the same informational-poster prompt; captions specify tier, actual dimensions and format:

A dense informational poster titled "CITY TRANSIT GUIDE" with three labeled columns headed "BUS", "METRO" and "TRAM", a short line of body text under each heading, clean Swiss grid layout, red and black ink on off-white paper, flat vector infographic style
LOW $0.043 MEDIUM $0.086 HIGH $0.3145

ideogram-v4.5/text-to-image, same prompt, one render each; all three are 720×1280 PNG. Left to right: LOW $0.043, MEDIUM $0.086, HIGH $0.3145.

Try this prompt →

The link carries only the prompt. Select portrait_16_9 and num_images=1 yourself, then choose LOW, MEDIUM, or HIGH for each run.

What you can see:

  • All three have the CITY TRANSIT GUIDE title and BUS, METRO, and TRAM columns.
  • LOW leaves more space through the center; MEDIUM adds route-like lines below the columns; HIGH has denser body blocks and bottom text.
  • Inspect characters, wording, and layout in each output. Paying for a tier does not guarantee every character is correct.

Editor price combinations

ideogram-v4.5/image-to-image bills per image too; regular and Precise (edit_precision=high) share tier prices, and VERY_LOW works with regular only:

Combination Per image Best for
regular · VERY_LOW $0.0115 Bulk draft edits
regular · LOW $0.043 Everyday edits
regular · MEDIUM (default) $0.086 Standard delivery
regular · HIGH $0.3145 Heavy repaints
Precise · LOW $0.043 Text and date fixes; inspect preserved areas
Precise · MEDIUM $0.086 Final retouching
Precise · HIGH $0.3145 High-detail retouching

A mask lowers the reference cap from four to three; prices don't move. Precise Edit usage is in the API guide.

Choosing tiers and spending less

  • Submit separate LOW tasks for alternatives: one image per task, four tasks for $0.172; vary the prompt or seed to explore compositions.
  • Start routine-delivery evaluation on MEDIUM: compare the default tier with LOW using the same prompt, then decide whether layout and detail fit the task.
  • Compare tiers for small-type finals: try HIGH when more detail is useful; still proofread the output rather than treating its tier as a guarantee.
  • Try Precise LOW for text fixes: a name or date change costs $0.043; inspect the edited and preserved areas before deciding whether to move up a tier.
  • Disable prompt expansion: pass enable_prompt_expansion: false to stop automatic expansion. Still inspect whether the output follows your instructions; the tier price stays unchanged.

Compared with related models (HiAPI prices)

Model Per image Notes
fibo-gen-1.5/text-to-image $0.0575 Photoreal + same-price editor
ideogram-v4.5/text-to-image (this model) LOW $0.043; MEDIUM $0.086; HIGH $0.3145 Typography-first
recraft-v4.1/text-to-image $0.05 Brand color parameters
gpt-image-2.5-flare/text-to-image 1K $0.025 Tiered resolution
krea-3/text-to-image $0.036 Prompt reasoning
gemini-nano-banana-2.1 1K $0.011 1K image tier

Monthly budgets (generation)

Images / month All LOW All MEDIUM All HIGH
100 $4.30 $8.60 $31.45
1,000 $43.00 $86.00 $314.50
10,000 $430.00 $860.00 $3,145.00

The HIGH gap is steep: 1,000 images cost $228.50 more than MEDIUM and $271.50 more than LOW. Explore on LOW, use MEDIUM for routine work, and try HIGH only for selected finals that need a closer small-type check.

FAQ

Does omitting quality select the lowest-priced tier?

No. The default is MEDIUM at $0.086 per image. Set quality: "low" explicitly for LOW at $0.043.

Can one request return several images?

No. Generation and editing currently return 1 image per task; num_images must be 1. Submit separate tasks for alternatives.

Can VERY_LOW be used with Precise Edit?

No. VERY_LOW costs $0.0115 and requires edit_precision: "regular". Precise starts at LOW for $0.043 per image.

Do size changes or references add a charge?

Size does not change the tier rate; editing references and masks do not add a fee. Up to 4 references are supported without a mask, or 3 with one.

How many LOW images can $10 cover?

At $0.043 each, $10 covers 232 LOW images for $9.976 in total. Submit one task per image.

Next: take the CITY TRANSIT GUIDE prompt and generate one LOW poster to check the layout; code and parameters are in the API guide. Related: generation and editing API guide · 4 editing prompt examples.

Related articles

View all articles →

Start building with persistent AI artifacts

Get your API key and call image, video and audio models through one task-based API.

Read the API docs