Model comparison
Kimi K3 vs Muse Spark 1.3
KimiCompare published specifications, vendor API list prices and measured Borg performance. Both models are included in Borg.
Kimi K3
Muse Spark 1.3
Included in Borg
At a glance
| Specification | Kimi K3 | Muse Spark 1.3 |
|---|---|---|
| Provider | Moonshot AI | Meta |
| Release | 2026-07-16 Vendor source | 2026-09-02 Vendor source |
| Context | 1,048,576 tokens Vendor source | 1,048,576 tokens Vendor source |
| Modalities | Input: text, image, video; output: Not published | Input: text, image, video, audio, PDF; output: text |
| Reasoning / thinking modes | Yes / Not published | Yes / Not published |
| Open weights | Yes | No |
| Vendor API list price / 1M tokens | $3 input / $15 output | $1.25 input / $4.25 output |
| Borg | Included in Borg subscription | Included in Borg subscription |
Moonshot AI
MetaMeasured on Borg
Metrics are not directly comparable to vendor benchmark claims. Models served different requests; this is observed gateway traffic, not a controlled head-to-head test.
Time to first token (gateway proxy)
Output tokens per second (whole turn)
| Metric | Kimi K3 | Muse Spark 1.3 |
|---|---|---|
| Time to first token (gateway proxy) p50 | 8627 ms (7633 samples) | 5938 ms (1912 samples) |
| Tokens per second (end to end) p50 | 18.1 output tokens per second (152924 samples) | 26.2 output tokens per second (1928 samples) |
| Request success rate | Not published | Not published |
| Coding bench pass rate | Not published | Not published |
Measurement window, samples and methodology
The fleet reference is the median of published model medians, not a request-weighted result. Prompt lengths, reasoning settings and traffic differ between models.
Kimi K3: Measured on Borg: SearchAtlas AI Gateway production request logs. 30-day window ending 2026-10-08; 183015 requests. Minimum 100 samples per published request metric.
Muse Spark 1.3: Measured on Borg: SearchAtlas AI Gateway production request logs. 30-day window ending 2026-10-08; 2340 requests. Minimum 100 samples per published request metric.
Completion speed covers the whole turn including queueing and reasoning time; it is not a streaming tokens-per-second figure.
- Time to first token (gateway proxy): time until the first streamed response event reaches the gateway; our closest available measure of time to first token
- Tokens per second (end to end): tokens per second, end to end: provider-reported output tokens divided by the whole turn's elapsed time, including queueing and reasoning; not a streaming per-token rate
How to choose
Kimi K3 is listed for long-horizon coding across large repositories, whereas Muse Spark 1.3 is listed for long-horizon coding with changing requirements. Choosing between Kimi K3 and Muse Spark 1.3 requires checking published specifications against observed Borg traffic for your workflow.
Key differences
| Spec | Kimi K3 | Muse Spark 1.3 |
|---|---|---|
| Context window | 1,048,576 tokens Vendor source | 1,048,576 tokens Vendor source |
| Reasoning modes | Yes / Not published | Yes / Not published |
| Release date | 2026-07-16 Vendor source | 2026-09-02 Vendor source |
| API price input / 1M | 3 | 1.25 Lower listed rate |
| API price output / 1M | 15 | 4.25 Lower listed rate |
| Time to first token (gateway proxy) | 8627 ms (7633 samples) | 5938 ms (1912 samples) Lower median |
| Output tokens per second (whole turn) | 18.1 output tokens/s (whole turn) (152924 samples) | 26.2 output tokens/s (whole turn) (1928 samples) Higher measured rate |
| Request success | Not published | Not published |
| Coding bench | Not published | Not published |
FAQ
How do Kimi K3 and Muse Spark 1.3 compare for coding?
Kimi K3 has a Borg coding benchmark of not published; Muse Spark 1.3 has not published. The Kimi K3–Muse Spark 1.3 coding comparison cannot establish an overall winner from this small test alone.
How quickly do Kimi K3 and Muse Spark 1.3 start responding on Borg?
Borg median first response for Kimi K3 is 8,627 ms; for Muse Spark 1.3 it is 5,938 ms. The Kimi K3–Muse Spark 1.3 gateway medians are observations, not guaranteed latency for individual requests.
What are the API prices for Kimi K3 and Muse Spark 1.3?
The listed API rate for Kimi K3 is $3 input / $15 output per 1M tokens; for Muse Spark 1.3 it is $1.25 input / $4.25 output per 1M tokens. Neither Kimi K3 nor Muse Spark 1.3 has its Borg subscription charge represented by those provider API rates. Kimi K3 vendor source · Muse Spark 1.3 vendor source
Claude