Side-by-side comparison of pricing, 12 benchmarks, and generation speed.
| Metric | LFM2.5-8B-A1B | GLM-4.5V (Non-reasoning) |
|---|---|---|
| Input ($/M tokens) | $0 | $0.6 |
| Output ($/M tokens) | $0 | $1.8 |
Verdict. LFM2.5-8B-A1B wins the overall benchmark matchup 1–0 across 1 overlapping categories, but raw benchmark score is only one input to the decision.
Pricing. Pricing varies significantly between these models — check the table above for the exact per-token rates. Many production workloads actually surface input-token cost (retrieval-augmented prompts, code-context windows), so factor both directions.
Strengths. LFM2.5-8B-A1B is strongest on Intelligence Index (8.1). GLM-4.5V (Non-reasoning) leads on Intelligence Index (6.8).
Speed. On throughput, LFM2.5-8B-A1B generates tokens at 339 tok/s versus 92 tok/s — about 73% faster. On time-to-first-token, GLM-4.5V (Non-reasoning) responds in 1850ms vs 1950ms, which matters most for chat-style UIs.
Provider. Liquid AI and Z AI sell to overlapping but distinct developer audiences: Liquid AI tends to ship frontier reasoning models with premium positioning, while Z AI often prices more aggressively. Your existing vendor relationships, billing, and SLA preferences may matter as much as the raw numbers above.
Workload cost. Workload scenarios (per million requests at 30M input + 15M output tokens): LFM2.5-8B-A1B costs $0.00 ($0/year); GLM-4.5V (Non-reasoning) costs $45.00 ($540/year). At a smaller 5M-input/2M-output scale (single-developer tool or prototype): LFM2.5-8B-A1B ≈ $0.00/run, GLM-4.5V (Non-reasoning) ≈ $6.60/run. At agent/realtime scale (200M input / 100M output per million requests): LFM2.5-8B-A1B ≈ $0/run, GLM-4.5V (Non-reasoning) ≈ $300/run. LFM2.5-8B-A1B becomes more attractive at higher volume — the absolute per-token pricing difference compounds when you ship at scale.
Recommendation. Both models have legitimate use cases — the right answer depends on whether you are optimizing for benchmark ceiling, latency, or unit cost. Start with the cheaper / faster model, evaluate against your specific task, and only switch if the upgrade shows a meaningful lift.
Head-to-head deltas
Data from Artificial Analysis API — 12 benchmarks
LFM2.5-8B-A1B is cheaper overall. Its blended price (3:1 input/output ratio) is $0.00/M tokens vs $0.90/M for GLM-4.5V (Non-reasoning).
LFM2.5-8B-A1B wins 1 out of 12 benchmarks compared to 0 for GLM-4.5V (Non-reasoning). See the detailed benchmark chart above for per-category results.
LFM2.5-8B-A1B generates tokens faster at 339 tok/s vs 92 tok/s. However, GLM-4.5V (Non-reasoning) has lower time-to-first-token (1.85s vs 1.95s).
Choose based on your priorities: LFM2.5-8B-A1B for lower cost, LFM2.5-8B-A1B for stronger benchmark performance, and LFM2.5-8B-A1B for faster generation. For latency-sensitive apps, check the TTFT comparison above.