Anthropic · in Borg
AnthropicClaude Haiku 4.5
ClaudeClaude Haiku 4.5 is the compact, fastest member of Anthropic’s current Claude lineup, released after Sonnet 4 as a cheaper way to run capable coding and tool-use tasks.
Reasoning200,000 tokenstext, image
Included in Borg
Download Borg Borg for BusinessAt a glance
| Specification | Value |
|---|---|
| Provider | Anthropic |
| Family | Claude |
| Release date | 2025-10-15 Vendor source |
| Context window | 200,000 tokens Vendor source |
| Input modalities | text, image |
| Output modalities | text |
| Reasoning | Yes |
| Open weights | No |
| Knowledge cutoff | 2025-02 Vendor source |
| Borg lane ids | claude-haiku-4-5 |
| Median time to first token (gateway proxy) on Borg | 889 ms |
| Median tokens per second (end to end) on Borg | 64.9 output tokens per second |
Not published by the vendor: thinking modes, fusion roles.
What Claude Haiku 4.5 is good for
Claude Haiku 4.5 is the compact, fastest member of Anthropic’s current Claude lineup, released after Sonnet 4 as a cheaper way to run capable coding and tool-use tasks. It supports extended thinking, unlike older Haiku models, but costs less per token and responds faster than the larger Sonnet, Opus, and Fable tiers. Its context and output limits are smaller than those of the newer flagship models, so it is better suited to bounded requests than huge, long-running agent sessions. It works well for quick code assistance, short document extraction, and many parallel classification or customer-support jobs; harder multistep reasoning can be escalated to a larger sibling. Borg includes Haiku 4.5 in the subscription, and Borg Fusion can route lighter work to it or you can select its lane directly for speed.
- high-volume classification and tagging
- quick bug explanations and small code edits
- short document extraction
- low-latency tool-driven assistants
How Borg uses it
Claude Haiku 4.5 is available as a subscription catalog lane in Borg. The catalog does not publish which Fusion profiles route to this model, so no routing role is claimed here. You can pick it directly by setting the OpenAI-compatible gateway model field to a listed lane id. Check the current catalog for availability before building a workflow around a lane.
OpenAImodel: "claude-haiku-4-5"Measured on Borg
| Metric | Median | 95th percentile | Samples |
|---|---|---|---|
| Time to first token (gateway proxy) | 889 ms | 2307 ms | 6197 |
| Tokens per second (end to end) | 64.9 output tokens per second | 88.6 output tokens per second | 226302 |
Time to first token (gateway proxy)
Output tokens per second (whole turn)
- Time to first token (gateway proxy): time until the first streamed response event reaches the gateway; our closest available measure of time to first token
- Tokens per second (end to end): tokens per second, end to end: provider-reported output tokens divided by the whole turn's elapsed time, including queueing and reasoning; not a streaming per-token rate
Completion speed covers the whole turn including queueing and reasoning time; it is not a streaming tokens-per-second figure.
Measured on Borg: SearchAtlas AI Gateway production request logs. 30-day window ending 2026-10-08; 289125 requests. Minimum 100 samples per published request metric.
Benchmarks
| Benchmark | Vendor-reported score | Source |
|---|---|---|
| SWE-bench Verified | 73.3% | Vendor source |
| Terminal-Bench | 41.0% | Vendor source |
| OSWorld | 50.7% | Vendor source |
| GPQA Diamond | 73.0% | Vendor source |
| MMMU (validation) | 73.2% | Vendor source |
Pricing
Included in Borg subscription. Vendor API list price: $1 input / $5 output per 1M tokens. Vendor price source.
Compare subscription and API pricing · Cost Savings Calculator
FAQ
What is Claude Haiku 4.5?
Claude Haiku 4.5 is listed in Borg's subscription catalog as claude-haiku-4-5. Claude Haiku 4.5 is the compact, fastest member of Anthropic’s current Claude lineup, released after Sonnet 4 as a cheaper way to run capable coding and tool-use tasks.
Is Claude Haiku 4.5 in Borg?
Yes. Claude Haiku 4.5 is listed in Borg's subscription catalog under the lane claude-haiku-4-5. You can choose the lane directly, or explore the available Fusion profiles. Availability and model details can change; check the current catalog before deployment.
What context window does Claude Haiku 4.5 have?
The published context window is 200,000 tokens. Consult the cited sources for details and current limits. Vendor source
Is Claude Haiku 4.5 good for coding?
Vendor-published evaluations appear in the benchmarks section; their settings may differ from your tasks. Its listed uses include high-volume classification and tagging, quick bug explanations and small code edits, short document extraction, low-latency tool-driven assistants. Try the lane for your own tasks, and compare the output with your requirements before choosing a default.
How fast is Claude Haiku 4.5 on Borg?
Measured on Borg: SearchAtlas AI Gateway production request logs. 30-day window ending 2026-10-08; 289125 requests. Minimum 100 samples per published request metric.
How much does Claude Haiku 4.5 cost on Borg vs the API?
This model is included in the Borg subscription. The vendor API list prices appear below with a source link. See the pricing explainer and cost savings calculator to compare billing approaches.
Sources
Related
Other Anthropic models: Claude Fable 5.1 · Claude Opus 5.5 · Claude Sonnet 5.5 · Claude Opus 4.6 · Claude Sonnet 4.6
Models sharing a Fusion role: Not published