UI-TARS 7B vs GPT-4o-mini
UI-TARS 7B costs 3.0x less per output token than GPT-4o-mini ($0.20 vs $0.60 per million). Updated October 22, 2026; 1 benchmarks have a result for both.
UI-TARS 7B or GPT-4o-mini: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick UI-TARS 7B for output tokens 3.0x cheaper ($0.20 vs $0.60 per million) and work measured by Kagi LLM Benchmark, where it leads.
- Shared results
- 1
- UI-TARS 7B only
- 0
- GPT-4o-mini only
- 28
- Like-for-like categories
- 1 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
No call
Only GPT-4o-mini has a AA Intelligence Index result, 6.7.
Directional only
Reasoning
GPQA Diamond
No call
Only GPT-4o-mini has a GPQA Diamond result, 44.2%.
Directional only
Math
AIME
No call
Only GPT-4o-mini has a AIME result, 11.5%.
Directional only
Knowledge
MMLU Pro
No call
Only GPT-4o-mini has a MMLU Pro result, 62.7%.
Directional only
Long documents
Listed context window
No call
Both take 128K tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
UI-TARS 7B
UI-TARS 7B $0.0002, GPT-4o-mini $0.0004 per turn.
List prices
At a glance
- Coding index
- UI-TARS 7B not benchmarked, GPT-4o-mini not benchmarked.
- Widest category
- Reasoning, 8.1 points apart on average over 1 shared benchmark.
- Widest single gap
- Kagi LLM Benchmark: UI-TARS 7B 20.7% against GPT-4o-mini 28.8%, 8.1 points apart on a 0 to 100 scale.
- Output price
- UI-TARS 7B $0.20 against GPT-4o-mini $0.60 per million output tokens, 3.0x apart.
- Context window
- Both take 128K tokens.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | UI-TARS 7B | Bars | GPT-4o-mini | Ties |
|---|---|---|---|---|
| Reasoning | 0 | 1 | ||
| All shared boards | 0 | 1 |
CodingDirectional onlySix of these boards feed the Vibe Coding Index4 boards, none shared
| Benchmark | UI-TARS 7B | GPT-4o-mini | Leader |
|---|---|---|---|
| AA Coding Index | Not measured | 11.4 | Only GPT-4o-mini measured |
| LiveCodeBench (Vals AI) | Not measured | 26.4% | Only GPT-4o-mini measured |
| LiveCodeBench (Artificial Analysis run) | Not measured | 23.4% | Only GPT-4o-mini measured |
| SciCode | Not measured | 1.5% | Only GPT-4o-mini measured |
| Neither model has a result on the other 35 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | UI-TARS 7B | GPT-4o-mini | Leader |
|---|---|---|---|
| AA Intelligence Index | Not measured | 6.7 | Only GPT-4o-mini measured |
| Epoch Capabilities Index | Not measured | 126.5 | Only GPT-4o-mini measured |
Agentic workNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | UI-TARS 7B | GPT-4o-mini | Leader |
|---|---|---|---|
| Neither model has a result on 41 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | UI-TARS 7B | GPT-4o-mini | Leader |
|---|---|---|---|
| GPQA Diamond | Not measured | 44.2% | Only GPT-4o-mini measured |
| MMMU Pro (Vals AI) | Not measured | 56.6% | Only GPT-4o-mini measured |
| GPQA Diamond (Artificial Analysis run) | Not measured | 42.6% | Only GPT-4o-mini measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 4.2% | Only GPT-4o-mini measured |
| ARC-AGI-2 | Not measured | 0.0% | Only GPT-4o-mini measured |
| Kagi LLM Benchmark | 20.7% | 28.8% | GPT-4o-mini |
| Neither model has a result on the other 16 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
| Benchmark | UI-TARS 7B | GPT-4o-mini | Leader |
|---|---|---|---|
| AIME | Not measured | 11.5% | Only GPT-4o-mini measured |
| MATH 500 (Vals AI) | Not measured | 72.6% | Only GPT-4o-mini measured |
| MGSM (Vals AI) | Not measured | 86.2% | Only GPT-4o-mini measured |
| AA Math Index | Not measured | 14.7 | Only GPT-4o-mini measured |
| AIME 2025 (Artificial Analysis run) | Not measured | 14.7% | Only GPT-4o-mini measured |
| GSM8K | Not measured | 91.3% | Only GPT-4o-mini measured |
| Neither model has a result on the other 7 boards here | |||
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
| Benchmark | UI-TARS 7B | GPT-4o-mini | Leader |
|---|---|---|---|
| MMLU Pro | Not measured | 62.7% | Only GPT-4o-mini measured |
| CorpFin (Vals AI) | Not measured | 45.5% | Only GPT-4o-mini measured |
| TaxEval (Vals AI) | Not measured | 60.5% | Only GPT-4o-mini measured |
| MortgageTax (Vals AI) | Not measured | 54.5% | Only GPT-4o-mini measured |
| MedQA (Vals AI) | Not measured | 72.4% | Only GPT-4o-mini measured |
| MMLU Pro (Artificial Analysis run) | Not measured | 64.8% | Only GPT-4o-mini measured |
| MMLU | Not measured | 81.8% | Only GPT-4o-mini measured |
| Neither model has a result on the other 11 boards here | |||
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | UI-TARS 7B | GPT-4o-mini | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralDirectional onlyPublic boards, shown as published, kept out of the index4 boards, none shared
| Benchmark | UI-TARS 7B | GPT-4o-mini | Leader |
|---|---|---|---|
| IFBench | Not measured | 31.0% | Only GPT-4o-mini measured |
| DTBench | Not measured | 54.4% | Only GPT-4o-mini measured |
| LMCA | Not measured | 10.4% | Only GPT-4o-mini measured |
| PIQA | Not measured | 88.7% | Only GPT-4o-mini measured |
| Neither model has a result on the other 42 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | UI-TARS 7B | GPT-4o-mini | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
UI-TARS 7B costs less
Repository review
50K input + 3K output tokens
UI-TARS 7B costs less
Agent loop
200K input + 10K output tokens
UI-TARS 7B costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | ByteDance UI-TARS 7B | OpenAI GPT-4o-mini |
|---|---|---|
| Maker | ByteDance | OpenAI |
| Weights | Open weights | No public weights listed |
| Released | July 22, 2025 | July 18, 2024 |
| Context window | 128K | 128K |
| Input $/Mtok | $0.10 | $0.15 |
| Output $/Mtok | $0.20 | $0.60 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is UI-TARS 7B better than GPT-4o-mini for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick UI-TARS 7B for output tokens 3.0x cheaper ($0.20 vs $0.60 per million) and work measured by Kagi LLM Benchmark, where it leads.
Which is cheaper, UI-TARS 7B or GPT-4o-mini?
UI-TARS 7B costs $0.10 per million input tokens and $0.20 per million output tokens; GPT-4o-mini costs $0.15 and $0.60. Cost per solved task is known for neither of them.
Which has the bigger context window, UI-TARS 7B or GPT-4o-mini?
Neither: both take 128K tokens.
How many benchmarks is the UI-TARS 7B vs GPT-4o-mini comparison based on?
1 benchmarks have a result for both models as of October 22, 2026. UI-TARS 7B's coding index rests on 0 independent benchmarks and GPT-4o-mini's on 0. Where only one model has a result, the page says so instead of naming a winner.