Ember-1 vs Qwen3.8 2.4T A95B
Qwen3.8 2.4T A95B costs 2.5x less per output token than Ember-1 ($6.00 vs $15.00 per million). Updated October 22, 2026; no benchmark has tested both models yet.
- Coding
- 52.6
- Overall
- No data
- Output $/M
- $15.00
- Coding
- No data
- Overall
- No data
- Output $/M
- $6.00
Ember-1 or Qwen3.8 2.4T A95B: which to pick
Only Ember-1 has a coding score, 52.6 on the Vibe Coding Index; Qwen3.8 2.4T A95B has none, so the two cannot be ranked on coding.
Pick Qwen3.8 2.4T A95B for output tokens 2.5x cheaper ($6.00 vs $15.00 per million).
No benchmark has tested both yet. Ember-1 has results on Terminal-Bench 4 (Vals AI), Vibe Code Bench (Vals AI) and Code migration (Vals AI); Qwen3.8 2.4T A95B on AA Coding Index, SciCode and AA Intelligence Index. Compare them on price and context instead.
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
Only Ember-1 has a coding score, so no call is made.
One side not measured
Intelligence
AA Intelligence Index
No call
Only Qwen3.8 2.4T A95B has a AA Intelligence Index result, 39.9.
Directional only
Agentic work
AA Agentic Index
No call
Only Qwen3.8 2.4T A95B has a AA Agentic Index result, 50.1.
Directional only
Long documents
Listed context window
No call
Both take 1.05M tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen3.8 2.4T A95B
Ember-1 $0.011, Qwen3.8 2.4T A95B $0.0050 per turn.
List prices
At a glance
- Coding index
- Ember-1 52.6 from 2 benchmarks, Qwen3.8 2.4T A95B not benchmarked.
- Widest category
- None yet: no category has been tested on both models.
- Widest single gap
- None yet: no benchmark has tested both models.
- Output price
- Ember-1 $15.00 against Qwen3.8 2.4T A95B $6.00 per million output tokens, 2.5x apart.
- Context window
- Both take 1.05M tokens.
Benchmark by benchmark
Every board either model has a result on, a category to a fold; bars are each result on a 0 to 100 scale. Only one model is measured on each, so no row names a leader.
CodingDirectional onlySix of these boards feed the Vibe Coding Index7 boards measured
| Benchmark | Ember-1 | Qwen3.8 2.4T A95B | Leader |
|---|---|---|---|
| AA Coding Index | Not measured | 71.9 | Only Qwen3.8 2.4T A95B measured |
| Terminal-Bench 4 (Vals AI) | 20.7% | Not measured | Only Ember-1 measured |
| Vibe Code Bench (Vals AI) | 83.9% | Not measured | Only Ember-1 measured |
| Code migration (Vals AI) | 32.5% | Not measured | Only Ember-1 measured |
| IOI (Vals AI) | 46.3% | Not measured | Only Ember-1 measured |
| CyberBench (Vals AI) | 69.5% | Not measured | Only Ember-1 measured |
| SciCode | Not measured | 54.1% | Only Qwen3.8 2.4T A95B measured |
| Neither model has a result on the other 36 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index1 board measured
| Benchmark | Ember-1 | Qwen3.8 2.4T A95B | Leader |
|---|---|---|---|
| AA Intelligence Index | Not measured | 39.9 | Only Qwen3.8 2.4T A95B measured |
| Neither model has a result on the other 1 board here | |||
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index9 boards measured
| Benchmark | Ember-1 | Qwen3.8 2.4T A95B | Leader |
|---|---|---|---|
| AA Agentic Index | Not measured | 50.1 | Only Qwen3.8 2.4T A95B measured |
| Finance Agent (Vals AI) | 51.8% | Not measured | Only Ember-1 measured |
| Tax Agent Bench (Vals AI) | 31.1% | Not measured | Only Ember-1 measured |
| Legal Research Bench (Vals AI) | 44.7% | Not measured | Only Ember-1 measured |
| Excel Modeling Benchmark (Vals AI) | 63.4% | Not measured | Only Ember-1 measured |
| BioMysteryBench (Vals AI) | 67.0% | Not measured | Only Ember-1 measured |
| Terminal-Bench-Science (Vals AI) | 5.7% | Not measured | Only Ember-1 measured |
| Vals Science Index | 31.8% | Not measured | Only Ember-1 measured |
| MCP Atlas | Not measured | 84.5% | Only Qwen3.8 2.4T A95B measured |
| Neither model has a result on the other 33 boards here | |||
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index3 boards measured
| Benchmark | Ember-1 | Qwen3.8 2.4T A95B | Leader |
|---|---|---|---|
| MysteryMechanism (Vals AI) | 22.5% | Not measured | Only Ember-1 measured |
| GPQA Diamond (Artificial Analysis run) | Not measured | 93.5% | Only Qwen3.8 2.4T A95B measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 42.4% | Only Qwen3.8 2.4T A95B measured |
| Neither model has a result on the other 20 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index1 board measured
| Benchmark | Ember-1 | Qwen3.8 2.4T A95B | Leader |
|---|---|---|---|
| MathArena | Not measured | 31.6% | Only Qwen3.8 2.4T A95B measured |
| Neither model has a result on the other 13 boards here | |||
GeneralDirectional onlyPublic boards, shown as published, kept out of the index2 boards measured
| Benchmark | Ember-1 | Qwen3.8 2.4T A95B | Leader |
|---|---|---|---|
| Public Benefits Bench (Vals AI) | 66.2% | Not measured | Only Ember-1 measured |
| Long context reasoning (AA-LCR) | Not measured | 80.3% | Only Qwen3.8 2.4T A95B measured |
| Neither model has a result on the other 46 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
Qwen3.8 2.4T A95B costs less
Repository review
50K input + 3K output tokens
Qwen3.8 2.4T A95B costs less
Agent loop
200K input + 10K output tokens
Qwen3.8 2.4T A95B 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 | Fireworks Ember-1 | Qwen3.8 2.4T A95B |
|---|---|---|
| Maker | Fireworks | Qwen |
| Weights | No public weights listed | Open weights |
| Released | September 24, 2026 | August 12, 2026 |
| Context window | 1.05M | 1.05M |
| Input $/Mtok | $3.00 | $2.00 |
| Output $/Mtok | $15.00 | $6.00 |
| Coding index | 52.6 | Not benchmarked |
| Independent measurements behind it | 2 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Ember-1 better than Qwen3.8 2.4T A95B for coding?
Only Ember-1 has a coding score, 52.6 on the Vibe Coding Index; Qwen3.8 2.4T A95B has none, so the two cannot be ranked on coding. Pick Qwen3.8 2.4T A95B for output tokens 2.5x cheaper ($6.00 vs $15.00 per million).
Which is cheaper, Ember-1 or Qwen3.8 2.4T A95B?
Ember-1 costs $3.00 per million input tokens and $15.00 per million output tokens; Qwen3.8 2.4T A95B costs $2.00 and $6.00. Cost per solved task is known for neither of them.
Which has the bigger context window, Ember-1 or Qwen3.8 2.4T A95B?
Neither: both take 1.05M tokens.
Has any benchmark tested both Ember-1 and Qwen3.8 2.4T A95B?
No benchmark has tested both yet. Ember-1 has results on Terminal-Bench 4 (Vals AI), Vibe Code Bench (Vals AI) and Code migration (Vals AI); Qwen3.8 2.4T A95B on AA Coding Index, SciCode and AA Intelligence Index. Compare them on price and context instead. Data as of October 22, 2026.