Model comparison
Claude Haiku 4.5 vs Gemini 3.8 Flash
Claude
GeminiCompare published specifications, vendor API list prices and measured Borg performance. Both models are included in Borg.
Claude Haiku 4.5
Gemini 3.8 Flash
Included in Borg
At a glance
| Specification | Claude Haiku 4.5 | Gemini 3.8 Flash |
|---|---|---|
| Provider | Anthropic | |
| Release | 2025-10-15 Vendor source | 2026-09-02 Vendor source |
| Context | 200,000 tokens Vendor source | 1,048,576 tokens Vendor source |
| Modalities | Input: text, image; output: text | Input: text, image, video, audio, pdf; output: text |
| Reasoning / thinking modes | Yes / Not published | Yes / Not published |
| Open weights | No | No |
| Vendor API list price / 1M tokens | $1 input / $5 output | $0.75 input / $3.75 output Introductory rates through December 31, 2026. From January 1, 2027: $1.50 input / $7.50 output / $0.15 cached input per 1M tokens. |
| Borg | Included in Borg subscription | Included in Borg subscription |
Anthropic
GoogleMeasured on Borg
Metrics are not directly comparable to vendor benchmark claims. Models served different requests; this is observed gateway traffic, not a controlled head-to-head test.
Time to first token (gateway proxy)
Output tokens per second (whole turn)
| Metric | Claude Haiku 4.5 | Gemini 3.8 Flash |
|---|---|---|
| Time to first token (gateway proxy) p50 | 889 ms (6197 samples) | 2510 ms (1147 samples) |
| Tokens per second (end to end) p50 | 64.9 output tokens per second (226302 samples) | 29.2 output tokens per second (85551 samples) |
| Request success rate | Not published | Not published |
| Coding bench pass rate | Not published | Not published |
Measurement window, samples and methodology
The fleet reference is the median of published model medians, not a request-weighted result. Prompt lengths, reasoning settings and traffic differ between models.
Claude Haiku 4.5: 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.
Gemini 3.8 Flash: Measured on Borg: SearchAtlas AI Gateway production request logs. 30-day window ending 2026-10-08; 92815 requests. Minimum 100 samples per published request metric.
Completion speed covers the whole turn including queueing and reasoning time; it is not a streaming tokens-per-second figure.
- 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
How to choose
Claude Haiku 4.5 is listed for high-volume classification and tagging, whereas Gemini 3.8 Flash is listed for long-horizon coding agents. Choosing between Claude Haiku 4.5 and Gemini 3.8 Flash requires checking published specifications against observed Borg traffic for your workflow.
Key differences
| Spec | Claude Haiku 4.5 | Gemini 3.8 Flash |
|---|---|---|
| Context window | 200,000 tokens Vendor source | 1,048,576 tokens Vendor source Larger context |
| Reasoning modes | Yes / Not published | Yes / Not published |
| Release date | 2025-10-15 Vendor source | 2026-09-02 Vendor source |
| API price input / 1M | 1 | 0.75 Introductory rates through December 31, 2026. From January 1, 2027: $1.50 input / $7.50 output / $0.15 cached input per 1M tokens. Lower listed rate |
| API price output / 1M | 5 | 3.75 Introductory rates through December 31, 2026. From January 1, 2027: $1.50 input / $7.50 output / $0.15 cached input per 1M tokens. Lower listed rate |
| Time to first token (gateway proxy) | 889 ms (6197 samples) Lower median | 2510 ms (1147 samples) |
| Output tokens per second (whole turn) | 64.9 output tokens/s (whole turn) (226302 samples) Higher measured rate | 29.2 output tokens/s (whole turn) (85551 samples) |
| Request success | Not published | Not published |
| Coding bench | Not published | Not published |
FAQ
How do Claude Haiku 4.5 and Gemini 3.8 Flash compare for coding?
Claude Haiku 4.5 has a Borg coding benchmark of not published; Gemini 3.8 Flash has not published. The Claude Haiku 4.5–Gemini 3.8 Flash coding comparison cannot establish an overall winner from this small test alone.
How quickly do Claude Haiku 4.5 and Gemini 3.8 Flash start responding on Borg?
Borg median first response for Claude Haiku 4.5 is 889 ms; for Gemini 3.8 Flash it is 2,510 ms. The Claude Haiku 4.5–Gemini 3.8 Flash gateway medians are observations, not guaranteed latency for individual requests.
What are the API prices for Claude Haiku 4.5 and Gemini 3.8 Flash?
The listed API rate for Claude Haiku 4.5 is $1 input / $5 output per 1M tokens; for Gemini 3.8 Flash it is $0.75 input / $3.75 output per 1M tokens. Neither Claude Haiku 4.5 nor Gemini 3.8 Flash has its Borg subscription charge represented by those provider API rates. Claude Haiku 4.5 vendor source · Gemini 3.8 Flash vendor source