Granite 4.0 H Micro vs Hy-MT2 1.8B
Side-by-side comparison of API pricing, specs, benchmarks, and capabilities
| Specification |
by IBM
|
by Tencent
|
|---|---|---|
| Overview | ||
| Status | Current budget Open weights | Current budget Open weights |
| Released | Oct 19, 2025 | May 21, 2026 |
| Pricing per million tokens | ||
| Input | $0.017/Mtok | $0.044/Mtok |
| Output | $0.112/Mtok | $0.177/Mtok |
| Blended avg | $0.041/Mtok | $0.077/Mtok |
| Price basis | — | Tencent's own price, quoted by the FP8 endpoint Tencent operates for this model. Tencent's RMB rate card carries the matching row — Hy-MT2-Lite at ¥0.3 in / ¥1.2 out per 1M tokens, the only row on that card that converts to these figures — at 6.78 CNY/USD. Tencent publishes no cached-input rate for this family. |
| Specifications | ||
| Context window | 128K tokens | 8K tokens |
| Parameters | Proprietary | 1.8B |
| Speed (TPS) | — | — |
| Modalities | ||
| Input |
text
|
text
|
| Benchmarks sources: IBM model card, IBM model card (5-shot) / public leaderboards | ||
| MMMLU | 55.19 vendor | — |
| BFCL | 57.56 vendor | — |
| HumanEval | 81 vendor | — |
| Providers | ||
| Available from |
IBM — $0.017/$0.112/Mtok
|
Tencent — $0.044/$0.177/Mtok
|
Cost at scale
1M tokens · 50/50 input/output| Volume | Granite 4.0 H Micro | Hy-MT2 1.8B | Savings |
|---|---|---|---|
| 1M tokens | $0.04 | $0.08 | $0.04 (51.8%) |
| 10M tokens | $0.41 | $0.77 | $0.36 (46.6%) |
| 100M tokens | $4.08 | $7.73 | $3.65 (47.2%) |
| 1000M tokens | $40.75 | $77.25 | $36.5 (47.2%) |
When to pick which
distilled from the pricing and spec data above- cost dominates: $0.041/Mtok blended vs $0.077 — 47.2% less on the same 50/50 token mix
- you need the longer context: 128K tokens vs 8K (15.6×)
- this pairing gives Hy-MT2 1.8B no edge on price, context, cache discount, or measured performance — pick it only for qualitative fit (ecosystem, compliance, model behavior on your prompts)
The cheaper option here — Granite 4.0 H Micro costs $0.041/Mtok blended on IBM.
Summary
Granite 4.0 H Micro by IBM costs $0.017/Mtok input and $0.112/Mtok output, with a 128K-token context window. It supports text input.
Hy-MT2 1.8B by Tencent costs $0.044/Mtok input and $0.177/Mtok output, with a 8K-token context window. It supports text input.
On a blended cost basis, Granite 4.0 H Micro is 47.2% cheaper than Hy-MT2 1.8B.It also has a larger context window.
Note: Pricing is per million tokens. Actual costs vary with usage patterns, prompt caching, and batch discounts. Always verify against official provider pricing pages.