Cheaper overall
Claude Haiku 5.5
16% lower blended rate
Cheaper on long prompts
GLM 5.3 Flash
above 100,000 input tokens
Higher benchmark scores
Claude Haiku 5.5
leads on 9 of 14 figures
Cheaper cached input
Claude Haiku 5.5
$0.01 vs $0.03, 67% less
Where each one wins
Claude Haiku 5.5
- Cheaper input tokens $0.10 vs $0.15, 33% less
- Cheaper cached input $0.01 vs $0.03, 67% less
- Higher overall intelligence (Intelligence Index) 43.4 vs 41.8
- Higher expert-exam performance (Humanity's Last Exam) 44.4 vs 39.9
- Higher long-context reasoning (AA-LCR) 82.7 vs 80
- Ahead on 6 more benchmark figures
- Offers Batch pricing not listed for GLM 5.3 Flash
GLM 5.3 Flash
- Cheaper on long prompts above 100,000 input tokens
- Higher real-world task performance (GDPval) 57.2 vs 56
- Higher answer reliability (Omniscience Non-Hallucination) 72.4 vs 59.6
- Higher Codecategories Elo 1,286 vs 1,280
- Ahead on 2 more benchmark figures
What four workloads cost
One run, and the same run a thousand times.
The gap opens on long prompts: GLM 5.3 Flash costs $0.0615 against $0.2075, a difference of $146.00 over a thousand runs.
| Workload | Claude Haiku 5.5 | GLM 5.3 Flash | ×1,000 | Difference |
|---|---|---|---|---|
| Chat turn 1,000 in / 500 out | $0.0004 | $0.0004 | $0.3500 / $0.4000 | Claude Haiku 5.5 is 13% cheaper |
| RAG answer 10,000 in / 800 out | $0.0014 | $0.0019 | $1.40 / $1.90 | Claude Haiku 5.5 is 26% cheaper |
| Agent loop 32,000 in / 700 out × 20 calls | $0.0170 | $0.0310 | $17.00 / $31.00 | Claude Haiku 5.5 is 45% cheaper |
| Long document 400,000 in / 3,000 out | $0.2075 | $0.0615 | $207.50 / $61.50 | GLM 5.3 Flash is 70% cheaper |
Price per 1M tokens
Above 100,000 input tokens the rates change, and the new rate applies to the whole request, not just the tokens past that point: Claude Haiku 5.5's input goes from $0.1000 to $0.5000. That is where the cheaper model changes.
| Context band | Price | Claude Haiku 5.5 | GLM 5.3 Flash | Difference |
|---|---|---|---|---|
| Prompts up to 100,000 tokens | Input | $0.10 | $0.15 | Claude Haiku 5.5 is 33% cheaper |
| Prompts up to 100,000 tokens | Cached input | $0.01 | $0.03 | Claude Haiku 5.5 is 67% cheaper |
| Prompts up to 100,000 tokens | Output | $0.50 | $0.50 | Same |
| Prompts over 100,000 tokens | Input | $0.50 | $0.15 | GLM 5.3 Flash is 70% cheaper |
| Prompts over 100,000 tokens | Cached input | $0.05 | $0.03 | GLM 5.3 Flash is 40% cheaper |
| Prompts over 100,000 tokens | Output | $2.50 | $0.50 | GLM 5.3 Flash is 80% cheaper |
Blended: Claude Haiku 5.5 $0.20, GLM 5.3 Flash $0.2375 per 1M — one rate at 3:1 input to output, for comparing two models at a glance.
Claude Haiku 5.5 priced by Anthropic, GLM 5.3 Flash by Z.ai. Standard tier, pay-as-you-go.
Other pricing tiers
Only Claude Haiku 5.5 lists Batch, at up to 50% off its own standard rate; GLM 5.3 Flash does not offer it.
Every tier either model sells is here, so this is the whole price sheet. A saving is measured against that model's own standard rate — how we read tiers.
| Tier | Claude Haiku 5.5 | GLM 5.3 Flash | Cheaper |
|---|---|---|---|
| Batch only on Claude Haiku 5.5 | $0.05 in / $0.25 out 50% off Standard | Not offered | — |
Benchmarks
They share 14 figures: Claude Haiku 5.5 leads on 9, GLM 5.3 Flash on 5. The widest gap is 37 points, on 3d Elo.
Overall intelligence
Intelligence Index — Overall intelligence across reasoning, knowledge & math evals
Claude Haiku 5.5 ahead by 1.6
Expert-exam performance
Humanity's Last Exam — Expert-level questions across many academic domains
Claude Haiku 5.5 ahead by 4.5
Long-context reasoning
AA-LCR — Long-context reasoning across large inputs
Claude Haiku 5.5 ahead by 2.7
Scientific coding
SciCode — Research-level scientific coding tasks
Claude Haiku 5.5 ahead by 3.4
Physics research reasoning
CritPt — Unpublished physics research reasoning problems
Claude Haiku 5.5 ahead by 3.5
Real-world task performance
GDPval — Economically valuable, real-world knowledge work
GLM 5.3 Flash ahead by 1.2
Factual accuracy
Omniscience Accuracy — Breadth of factual knowledge across domains
Claude Haiku 5.5 ahead by 8.9
Answer reliability
Omniscience Non-Hallucination — How reliably the model avoids fabricated answers
GLM 5.3 Flash ahead by 12.8
3d Elo
Design Arena head-to-head rating in the 3d category
Claude Haiku 5.5 ahead by 37
Codecategories Elo
Design Arena head-to-head rating in the codecategories category
GLM 5.3 Flash ahead by 6
Dataviz Elo
Design Arena head-to-head rating in the dataviz category
Claude Haiku 5.5 ahead by 28
Gamedev Elo
Design Arena head-to-head rating in the gamedev category
Claude Haiku 5.5 ahead by 24
Uicomponent Elo
Design Arena head-to-head rating in the uicomponent category
GLM 5.3 Flash ahead by 13
Website Elo
Design Arena head-to-head rating in the website category
GLM 5.3 Flash ahead by 8
Each figure is that model's best published run, with the effort level named beside it — where these scores come from.
Specifications
GLM 5.3 Flash also takes video. Only GLM 5.3 Flash lists open weights.
| Specification | Claude Haiku 5.5 | GLM 5.3 Flash |
|---|---|---|
| Context window | 1,000,000 tokens | 1,000,000 tokens |
| Max output | 128,000 tokens | 128,000 tokens |
| Takes in | Text, image, file | Text, image, video, file |
| Puts out | Text | Text |
| Knowledge cutoff | 30 Jun 2026 | — |
| Released | 7 Oct 2026 | 26 Aug 2026 |
| Status | Active | Active |
| Sold by | Anthropic, Perplexity | Perplexity, Z.ai |
| Tool use | Yes | Yes |
| Structured outputs | Yes | Yes |
| Web search | Not listed | Not listed |
| Prompt caching | Yes | Yes |
| Code execution | Yes | Yes |
| Computer use | Yes | Yes |
| Open weights | Not listed | Yes |
| Reasoning | Yes | Yes |
FAQs
- Is Claude Haiku 5.5 cheaper than GLM 5.3 Flash?
- Yes. A 1,000-token prompt with a 500-token reply costs $0.0004 on Claude Haiku 5.5 and $0.0004 on GLM 5.3 Flash.
- How much do Claude Haiku 5.5 and GLM 5.3 Flash cost per 1M tokens?
- Claude Haiku 5.5 costs $0.10 for input and $0.50 for output. GLM 5.3 Flash costs $0.15 and $0.50. Standard pay-as-you-go rates, as of 7 Oct 2026.
- Which is better, Claude Haiku 5.5 or GLM 5.3 Flash?
- On published benchmarks, Claude Haiku 5.5. It leads on 9 of the 14 figures both models report, including overall intelligence (Intelligence Index), where it scores 43.4 against 41.8.
- Is Claude Haiku 5.5 or GLM 5.3 Flash cheaper for long prompts?
- GLM 5.3 Flash, above 100,000 input tokens. A rate that changes at a context length applies to the whole request, not only the tokens past it, so the switch is sharp rather than gradual.
- Does Claude Haiku 5.5 or GLM 5.3 Flash offer batch pricing?
- Claude Haiku 5.5 only. Its batch input costs $0.05, 50% off its standard rate, and GLM 5.3 Flash lists no batch tier at all.