Nova Premier 1.0 vs UI-TARS 7B
UI-TARS 7B costs 62.5x less per output token than Nova Premier 1.0 ($0.20 vs $12.50 per million). Updated October 22, 2026; 1 benchmarks have a result for both.
Nova Premier 1.0 or UI-TARS 7B: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Nova Premier 1.0 for a larger context window (1M against 128K tokens).
- Shared results
- 1
- Nova Premier 1.0 only
- 13
- UI-TARS 7B only
- 0
- 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 Nova Premier 1.0 has a AA Intelligence Index result, 9.2.
Directional only
Web design
Design Arena: websites
No call
Only Nova Premier 1.0 has a Design Arena: websites result, 835.
Directional only
Long documents
Listed context window
Nova Premier 1.0
Nova Premier 1.0 takes 1M tokens against 128K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
UI-TARS 7B
Nova Premier 1.0 $0.0088, UI-TARS 7B $0.0002 per turn.
List prices
At a glance
- Coding index
- Nova Premier 1.0 not benchmarked, UI-TARS 7B not benchmarked.
- Widest category
- Reasoning, 24.1 points apart on average over 1 shared benchmark.
- Widest single gap
- Kagi LLM Benchmark: Nova Premier 1.0 44.8% against UI-TARS 7B 20.7%, 24.1 points apart on a 0 to 100 scale.
- Output price
- Nova Premier 1.0 $12.50 against UI-TARS 7B $0.20 per million output tokens, 62.5x apart.
- Context window
- Nova Premier 1.0 takes 1M tokens, UI-TARS 7B 128K.
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 | Nova Premier 1.0 | Bars | UI-TARS 7B | Ties |
|---|---|---|---|---|
| Reasoning | 1 | 0 | ||
| All shared boards | 1 | 0 |
CodingDirectional onlySix of these boards feed the Vibe Coding Index3 boards, none shared
| Benchmark | Nova Premier 1.0 | UI-TARS 7B | Leader |
|---|---|---|---|
| LiveCodeBench (Artificial Analysis run) | 31.7% | Not measured | Only Nova Premier 1.0 measured |
| Terminal-Bench Hard (Artificial Analysis run) | 6.8% | Not measured | Only Nova Premier 1.0 measured |
| ALE-Bench | 147.4 | Not measured | Only Nova Premier 1.0 measured |
| Neither model has a result on the other 36 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
| Benchmark | Nova Premier 1.0 | UI-TARS 7B | Leader |
|---|---|---|---|
| AA Intelligence Index | 9.2 | Not measured | Only Nova Premier 1.0 measured |
| Neither model has a result on the other 1 board here | |||
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
| Benchmark | Nova Premier 1.0 | UI-TARS 7B | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 38.3% | Not measured | Only Nova Premier 1.0 measured |
| Neither model has a result on the other 40 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Nova Premier 1.0 | UI-TARS 7B | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | 56.9% | Not measured | Only Nova Premier 1.0 measured |
| Humanity's Last Exam (Artificial Analysis run) | 4.2% | Not measured | Only Nova Premier 1.0 measured |
| Kagi LLM Benchmark | 44.8% | 20.7% | Nova Premier 1.0 |
| Neither model has a result on the other 19 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Nova Premier 1.0 | UI-TARS 7B | Leader |
|---|---|---|---|
| AA Math Index | 17.3 | Not measured | Only Nova Premier 1.0 measured |
| AIME 2025 (Artificial Analysis run) | 17.3% | Not measured | Only Nova Premier 1.0 measured |
| Neither model has a result on the other 11 boards here | |||
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
| Benchmark | Nova Premier 1.0 | UI-TARS 7B | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | 73.3% | Not measured | Only Nova Premier 1.0 measured |
| Neither model has a result on the other 17 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
| Benchmark | Nova Premier 1.0 | UI-TARS 7B | Leader |
|---|---|---|---|
| Design Arena: websites | 835 | Not measured | Only Nova Premier 1.0 measured |
| Neither model has a result on the other 8 boards here | |||
GeneralDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Nova Premier 1.0 | UI-TARS 7B | Leader |
|---|---|---|---|
| IFBench | 36.2% | Not measured | Only Nova Premier 1.0 measured |
| Long context reasoning (AA-LCR) | 34.3% | Not measured | Only Nova Premier 1.0 measured |
| Neither model has a result on the other 44 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Nova Premier 1.0 | UI-TARS 7B | 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 | Amazon Nova Premier 1.0 | ByteDance UI-TARS 7B |
|---|---|---|
| Maker | Amazon | ByteDance |
| Weights | No public weights listed | Open weights |
| Released | October 31, 2025 | July 22, 2025 |
| Context window | 1M | 128K |
| Input $/Mtok | $2.50 | $0.10 |
| Output $/Mtok | $12.50 | $0.20 |
| 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 Nova Premier 1.0 better than UI-TARS 7B for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Nova Premier 1.0 for a larger context window (1M against 128K tokens).
Which is cheaper, Nova Premier 1.0 or UI-TARS 7B?
Nova Premier 1.0 costs $2.50 per million input tokens and $12.50 per million output tokens; UI-TARS 7B costs $0.10 and $0.20. Cost per solved task is known for neither of them.
Which has the bigger context window, Nova Premier 1.0 or UI-TARS 7B?
Nova Premier 1.0, with 1M tokens against 128K.
How many benchmarks is the Nova Premier 1.0 vs UI-TARS 7B comparison based on?
1 benchmarks have a result for both models as of October 22, 2026. Nova Premier 1.0's coding index rests on 0 independent benchmarks and UI-TARS 7B's on 0. Where only one model has a result, the page says so instead of naming a winner.