← All models
Flash — fastest and cheapest, lighter reasoningOpen Weights
UI-TARS 7B Pricing
ByteDance · bytedance-ui-tars-1-5-7b
Input / 1M tokens
$0.100
Output / 1M tokens
$0.200
Context window
128k tokens
≈ 171 pages of text
Max output
2,048 tokens
What it costs in practice
Typical request (1,200 in + 400 out tokens)$0.0002
1,000 requests / month$0.2/mo10,000 requests / month$2.00/mo100,000 requests / month$20.00/moEstimate your own workloadOpens the calculator with UI-TARS 7B preloaded — adjust volume and token counts there.
Price history
Only one price point recorded so far.
The staircase chart appears once a price change is detected.
What it excels at
Performs basic multimodal tasks on UI screenshots and short documents at very low cost. Supports 128k context for retrieval-style inputs.
The business tradeoff
7B scale limits complex reasoning and instruction following compared with larger models. Output length capped at 2k tokens restricts long-form generation.
Head-to-head
Vendor list rates, as of Jun 15, 2026 · source: openrouter · per-request examples assume 1,200 input + 400 output tokens.