MiniCPM5-2B vs Muse Glimmer 30B vs Solar Pro 4
Too close to call on our weighted score (Muse Glimmer 30B 57, Solar Pro 4 55, MiniCPM5-2B 53). The right pick depends on what you value most.
OpenBMB
MiniCPM5-2B
53/100- ECI—
- Price$0.124 / $0.743
- Context131K
MetaMuse Glimmer 30B
57/100- ECI—
- Price$0.30 / $1.20
- Context131K
Upstage
Solar Pro 4
55/100- ECI—
- Price$0.30 / $1.20
- Context524K
Too close to call
It is close. Our weighted score puts them within 3 points (Muse Glimmer 30B 57/100, Solar Pro 4 55/100, MiniCPM5-2B 53/100), so choose by what matters most for your work: MiniCPM5-2B on price and Solar Pro 4 for long inputs. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
- CapabilityNot enough dataNo independent benchmark covers every model here yet
- Lowest priceMiniCPM5-2BMiniCPM5-2B $0.279 · Muse Glimmer 30B $0.525 · Solar Pro 4 $0.525 per 1M tokens (3:1 blend)
- Longest contextSolar Pro 4Solar Pro 4 524,288 · MiniCPM5-2B 131,072 · Muse Glimmer 30B 131,072 tokens
- Widest inputsMuse Glimmer 30BMiniCPM5-2B: Text · Muse Glimmer 30B: Text, Images · Solar Pro 4: Text
- Self-hostingMiniCPM5-2B and Muse Glimmer 30BPublishes downloadable weights (apache-2.0 and Apache 2.0)
| Measure | Weight | MiniCPM5-2B | Muse Glimmer 30B | Solar Pro 4 |
|---|---|---|---|---|
| Price | 50% | 76 | 63 | 63 |
| Inputs & features | 30% | 35 | 70 | 45 |
| Context window | 20% | 24 | 24 | 49 |
| Overall | 100% | 53/100 | 57/100 | 55/100 |
Left out because at least one model lacks the data: capability. The remaining weights were rescaled.
Every spec in one table
Highlighted cells lead their row. Dashes mean the data is not published.
| Specification | MiniCPM5-2BOpenBMB | Muse Glimmer 30BMeta | Solar Pro 4Upstage |
|---|---|---|---|
| Capability | |||
| Capabilities Index (ECI) | — | — | — |
| ECI rank | — | — | — |
| Price per million tokens | |||
| Input | $0.124 (best) | $0.30 | $0.30 |
| Output | $0.743 (best) | $1.20 | $1.20 |
| Cached input | — | — | $0.06 |
| Blended (3:1) | $0.279 (best) | $0.525 | $0.525 |
| Long-context rate | Same rate | Same rate | Same rate |
| Price source | Median of 1 providers | Median of 10 providers | Official Upstage API |
| Limits | |||
| Context window | 131,072 tokens | 131,072 tokens | 524,288 tokens (best) |
| Max output | 131,072 tokens | 131,072 tokens | 131,072 tokens |
| Inputs and features | |||
| Text | Yes | Yes | Yes |
| Images | No | Yes | No |
| PDFs | No | No | No |
| Audio | No | No | No |
| Video | No | No | No |
| Reasoning | Yes | Yes | Yesminimal · low · medium · high · xhigh · max |
| Tool calling | Yes | Yes | Yes |
| Structured output | No | Yes | Yes |
| Availability | |||
| Weights | Openapache-2.0 | OpenApache 2.0 | Proprietary |
| API model ID | — | — | solar-pro4 |
| API providers | 1 | 12 (best) | 3 |
| Released | Sep 6, 2026 | Aug 10, 2026 | Aug 6, 2026 |
| Knowledge cutoff | — | Jan 4, 2026 | Feb 2026 |
What would a month cost?
Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.
- MiniCPM5-2B$2.73
- Muse Glimmer 30B$5.40
- Solar Pro 4$5.40
Which should you choose?
Which is better: MiniCPM5-2B, Muse Glimmer 30B or Solar Pro 4?
It is close. Our weighted score puts them within 3 points (Muse Glimmer 30B 57/100, Solar Pro 4 55/100, MiniCPM5-2B 53/100), so choose by what matters most for your work: MiniCPM5-2B on price and Solar Pro 4 for long inputs. The score weighs price 50%, inputs & features 30%, context window 20%. None of the shared benchmarks cover every model here yet, so capability is left out of this verdict.
Which is cheaper, MiniCPM5-2B, Muse Glimmer 30B or Solar Pro 4?
MiniCPM5-2B is cheaper at $0.124 input / $0.743 output per million tokens (median across 1 API provider). Muse Glimmer 30B costs $0.30 input / $1.20 output per million tokens (median across 10 API providers); Solar Pro 4 costs $0.30 input / $1.20 output per million tokens (official Upstage API price). At a typical mix of three input tokens to one output token, that is $0.279 per million tokens for MiniCPM5-2B versus $0.525 for Muse Glimmer 30B (1.9× as much) and $0.525 for Solar Pro 4 (1.9× as much).
Which scores higher on benchmarks?
There is no independent benchmark that covers all three models yet. MiniCPM5-2B has not been scored yet, Muse Glimmer 30B has not been scored yet and Solar Pro 4 has not been scored yet.
Which is better for coding?
There are no published SWE-bench Verified results for MiniCPM5-2B, Muse Glimmer 30B and Solar Pro 4 yet, so there is no like-for-like coding score. Test both on a sample of your own tasks. All three support tool calling for agent workflows.
Which has the bigger context window?
Solar Pro 4 has the largest context window at 524,288 tokens, against 131,072 for MiniCPM5-2B and 131,072 for Muse Glimmer 30B. Maximum output per response: MiniCPM5-2B up to 131,072, Muse Glimmer 30B up to 131,072, Solar Pro 4 up to 131,072 tokens.
Which can read images, PDFs, audio or video?
MiniCPM5-2B accepts text; Muse Glimmer 30B accepts text and images; Solar Pro 4 accepts text. Muse Glimmer 30B handles the widest range of inputs.
Are any of these open source?
MiniCPM5-2B and Muse Glimmer 30B publishes its weights (apache-2.0 and Apache 2.0) and can be self-hosted; Solar Pro 4 is proprietary.
Which is newer?
MiniCPM5-2B is the newest, released Sep 6, 2026. Muse Glimmer 30B came out Aug 10, 2026; Solar Pro 4 came out Aug 6, 2026. Knowledge cutoff: Muse Glimmer 30B Jan 4, 2026, Solar Pro 4 Feb 2026.
How do you decide the winner?
Each model gets a 0–100 score on 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 Oct 4, 2026.
