GLM-5.3 vs Kimi K2.6 vs Muse Spark Open
Muse Spark Open comes out ahead, 68 to 65 and 63 on our weighted score, though Kimi K2.6 is 14% cheaper per token.
Price in rating: Included · Exclude price
Z.ai (Zhipu)
GLM-5.3
63/100- ECI157.1
- Price$1.40 / $4.40
- Context1M
Moonshot AI
Kimi K2.6
65/100- ECI151.6
- Price$0.95 / $4.00
- Context262K
- Our pick
Meta
Muse Spark Open
68/100- ECI155.5
- Price$1.25 / $4.25
- Context1.05M
Muse Spark Open is our pick
Muse Spark Open is the better all-round choice, scoring 68/100 against Kimi K2.6 (65) and GLM-5.3 (63). Kimi K2.6 wins on price. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
- CapabilityGLM-5.3Capabilities Index (ECI): GLM-5.3 157.1 · Muse Spark Open 155.5 · Kimi K2.6 151.6
- Lowest priceKimi K2.6Kimi K2.6 $1.71 · Muse Spark Open $2.00 · GLM-5.3 $2.15 per 1M tokens (3:1 blend)
- Longest contextMuse Spark OpenMuse Spark Open 1,048,576 · GLM-5.3 1,000,000 · Kimi K2.6 262,144 tokens
- Widest inputsMuse Spark OpenGLM-5.3: Text · Kimi K2.6: Text, Images, Video · Muse Spark Open: Text, Images, Audio, Video
- Self-hostingAll open weightsEvery model here can be downloaded and run on your own hardware
| Measure | Weight | GLM-5.3 | Kimi K2.6 | Muse Spark Open |
|---|---|---|---|---|
| CapabilityCapabilities Index (ECI) | 50% | 84 | 79 | 82 |
| Price | 25% | 33 | 37 | 34 |
| Inputs & features | 15% | 45 | 80 | 80 |
| Context window | 10% | 64 | 43 | 65 |
| Overall | 100% | 63/100 | 65/100 | 68/100 |
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | |||
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | 157.1 (best) | 151.6 | 155.5 |
| ECI rank | #33 of 158 (best) | #61 of 158 | #43 of 158 |
| GPQA DiamondGraduate-level science questions | 90.9% (best) | 90.8% | 88.6% |
| FrontierMath Tiers 1–3Research-level mathematics | 68.8% | 57.2% | 69.8% (best) |
| OTIS Mock AIME 2024–2025Competition mathematics | 91.1% | 96.1% | 98.1% (best) |
| SWE-bench VerifiedFixing real GitHub issues | — | 76.7% | — |
| SimpleQA VerifiedShort factual questions | 41.0% | 34.9% | 56.6% (best) |
| Price per million tokens | |||
| Input | $1.40 | $0.95 (best) | $1.25 |
| Output | $4.40 | $4.00 (best) | $4.25 |
| Cached input | $0.26 | $0.16 (best) | — |
| Blended (3:1) | $2.15 | $1.71 (best) | $2.00 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Official Z.AI API | Official Moonshot AI API | Median of 22 providers |
| Limits | |||
| Context window | 1,000,000 tokens | 262,144 tokens | 1,048,576 tokens (best) |
| Max output | 131,072 tokens | 262,144 tokens (best) | 131,072 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | Yes |
| PDFs | No | No | No |
| Audio | No | No | Yes |
| Video | No | Yes | Yes |
| Reasoning | Yeslow · high · max | Yes | Yes |
| Tool calling | Yes | Yes | Yes |
| Structured output | Yes | Yes | No |
| Availability | |||
| Weights | Open | Open | OpenMuse Model License |
| API model ID | glm-5.3 | kimi-k2.6 | — |
| API providers | 123 (best) | 64 | 22 |
| Released | Aug 14, 2026 | Apr 20, 2026 | Oct 15, 2026 |
| Knowledge cutoff | — | Jan 2025 | Jan 4, 2026 |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
GLM-5.3$22.80
Kimi K2.6$17.50
Muse Spark Open$21.00
Which should you choose?
Which is better: GLM-5.3, Kimi K2.6 or Muse Spark Open?
Muse Spark Open is the better all-round choice, scoring 68/100 against Kimi K2.6 (65) and GLM-5.3 (63). Kimi K2.6 wins on price. The score weighs capability 50%, price 25%, inputs & features 15%, context window 10%.
Which is cheaper, GLM-5.3, Kimi K2.6 or Muse Spark Open?
Kimi K2.6 is cheaper at $0.95 input / $4.00 output per million tokens (official Moonshot AI API price). Muse Spark Open costs $1.25 input / $4.25 output per million tokens (median across 22 API providers); GLM-5.3 costs $1.40 input / $4.40 output per million tokens (official Z.AI API price). At a typical mix of three input tokens to one output token, that is $1.71 per million tokens for Kimi K2.6 versus $2.00 for Muse Spark Open (1.2× as much) and $2.15 for GLM-5.3 (1.3× as much).
Which scores higher on benchmarks?
GLM-5.3 scores higher on the Capabilities Index (ECI): GLM-5.3 157.1 (#33 of 158), Muse Spark Open 155.5 (#43 of 158) and Kimi K2.6 151.6 (#61 of 158). The confidence ranges of the top two overlap (155.0–159.3 vs 153.4–157.7), so treat the gap as small. On individual benchmarks: GPQA Diamond — GLM-5.3 90.9%, Kimi K2.6 90.8%, Muse Spark Open 88.6%; FrontierMath Tiers 1–3 — Muse Spark Open 69.8%, GLM-5.3 68.8%, Kimi K2.6 57.2%; OTIS Mock AIME 2024–2025 — Muse Spark Open 98.1%, Kimi K2.6 96.1%, GLM-5.3 91.1%; SimpleQA Verified — Muse Spark Open 56.6%, GLM-5.3 41.0%, Kimi K2.6 34.9%.
Which is better for coding?
There are no published SWE-bench Verified results for GLM-5.3 and Muse Spark Open yet, so there is no like-for-like coding score. On overall capability, GLM-5.3 leads, which tends to carry over to coding, but test on your own codebase. All three support tool calling for agent workflows.
Which has the bigger context window?
Muse Spark Open has the largest context window at 1,048,576 tokens, against 1,000,000 for GLM-5.3 and 262,144 for Kimi K2.6. Maximum output per response: GLM-5.3 up to 131,072, Kimi K2.6 up to 262,144, Muse Spark Open up to 131,072 tokens.
Which can read images, PDFs, audio or video?
GLM-5.3 accepts text; Kimi K2.6 accepts text, images and video; Muse Spark Open accepts text, images, audio and video. Muse Spark Open handles the widest range of inputs.
Are any of these open source?
Yes, all three publish their weights (Muse Model License), so you can self-host them.
Which is newer?
Muse Spark Open is the newest, released Oct 15, 2026. GLM-5.3 came out Aug 14, 2026; Kimi K2.6 came out Apr 20, 2026. Knowledge cutoff: Kimi K2.6 Jan 2025, Muse Spark Open Jan 4, 2026.
How do you decide the winner?
This rating uses capability (50%, independent benchmark results); price (25%, blended price per million tokens (3 input : 1 output), log scale); inputs & features (15%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (10%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 18, 2026.