GPT 4.5 preview

OpenAI chat

GPT 4.5 preview is an OpenAI chat model.It supports a 128,000-token context windowwith up to 16,384 output tokens.Input is priced at $75.00/M tokens and output at $150/M tokens. Capabilities include function calling, vision, prompt caching. Route GPT 4.5 preview via Future AGI's Agent Command Center for unified observability, caching, and 15 routing strategies including cost-optimized fallback.

Pricing source: litellm Last verified: May 7, 2026 View source ↗
Cost calculator

Estimate GPT 4.5 preview spend

Pick a workload, fine-tune the sliders, and see the monthly bill.

~3K in / ~400 out · 5K req/day
3,000
0128,000
400
016,384
5,000
01,000,000
cached @ $37.50/M
Per request
$0.2850
in $0.2250 · out $0.0600
Per day
$1,425
5,000 requests
Per month
$43,373
152,188 requests

Estimate uses $75.00/M input · $150/M output. Provider pricing changes. Production costs vary with retries, streaming overhead, and tool-call rounds.
Want this for free? Cache + route via Agent Command Center — first 100K requests and 100K cache hits free every month.

Cheaper at a similar Arena tier

GPT 5.3 Chat latest (OpenAI) sits in the same Chatbot Arena tier (±80 ELO of GPT 4.5 preview) and runs ~96% cheaper for a typical RAG workload — $1,651/mo vs $43,373/mo at 3K in / 400 out · 5K reqs/day.

Quality match is gated on real benchmarks; the CTA disappears when no comparable peer exists.

Compare side-by-side →

Pricing

Per-token rates, expressed in USD per 1M tokens. Verified May 7, 2026.

Input $75.00/M
Output $150/M
Cached input $37.50/M
Batch input $37.50/M
Batch output $75.00/M

Limits

Context window
128,000 tokens
Max input
128,000 tokens
Max output
16,384 tokens
Modalities
vision, pdf, text

Capabilities

  • Function calling ✓ supported
  • Parallel tool calls ✓ supported
  • Vision input ✓ supported
  • Audio input — not advertised
  • Audio output — not advertised
  • PDF input ✓ supported
  • Streaming ✓ supported
  • Structured output ✓ supported
  • Prompt caching ✓ supported
  • Reasoning — not advertised

Where it's strong

  • +pricing — cheaper than 99% of priced chat models on Future AGI
  • +PDF input — only 16% of chat models on Future AGI advertise this
  • +parallel tool calls — only 21% of chat models on Future AGI advertise this
  • +prompt caching — only 23% of chat models on Future AGI advertise this

Watch out for

  • No major caveats flagged from public spec.

Benchmark scores

Reported public benchmark numbers. Each row links to the source.

Chatbot Arena ELOgeneral· overall
Captured May 12, 2026
Captured May 12, 2026
Captured May 12, 2026
Captured May 12, 2026
SWE-bench Verifiedagent· best internal performance
Captured May 12, 2026
Captured May 12, 2026
Try it

Call GPT 4.5 preview via Agent Command Center

One OpenAI-compatible endpoint. Routing, fallback, semantic caching, guardrails, and cost tracking come along for the ride. First 100K requests + 100K cache hits free every month.

SDK
Native Future AGI client (agentcc / @agentcc/client). Per-call metadata — provider, cost, latency, cache hit, request id — is returned on x-agentcc-* response headers, so any HTTP client can read it.
# GPT 4.5 preview via the Agent Command Center Python SDK
# pip install agentcc
import os
from agentcc import AgentCC

client = AgentCC(
    api_key=os.environ["AGENTCC_API_KEY"],   # from app.futureagi.com → Settings → API Keys
    base_url="https://gateway.futureagi.com/v1",
)

resp = client.chat.completions.create(
    model="openai/gpt-4-5-preview",
    messages=[{"role": "user", "content": "Hello, GPT 4.5 preview!"}],
)

print(resp.choices[0].message.content)
print(f"Tokens: {resp.usage.total_tokens}")

# Per-call gateway metadata is returned on x-agentcc-* response headers.
# When you need it programmatically, use .with_raw_response to get them:
raw = client.chat.completions.with_raw_response.create(
    model="openai/gpt-4-5-preview",
    messages=[{"role": "user", "content": "Same call, but I want the headers."}],
)
print("Provider:", raw.headers.get("x-agentcc-provider"))
print("Latency:", raw.headers.get("x-agentcc-latency-ms"), "ms")
print("Cost:   ", raw.headers.get("x-agentcc-cost"), "USD")
print("Cache:  ", raw.headers.get("x-agentcc-cache"))
Set AGENTCC_API_KEY with a key fromapp.futureagi.com.Gateway docs ↗
Advanced: fallback + cache config (YAML)
strategy: cost-optimized
targets:
  - model: gpt-4-5-preview
    provider: openai
    weight: 80
fallbacks:
  - model: gpt-5-3-chat-latest
    provider: openai
  - model: claude-opus-4-6-20260205
    provider: anthropic
guardrails: [pii, prompt-injection, secrets]
cache: { exact: true, semantic: true }

Same model on other providers

gpt-4-5-preview is also available via 1 other route. Pricing, regions, and capabilities can differ — compare before routing production traffic.

ProviderInput / 1MOutput / 1MVerified
Azure OpenAI$75.00/M$150/MMay 12, 2026

Compare with similar models

Grouped by Chatbot Arena tier (GPT 4.5 preview sits at 1444 ELO).

FAQ

How much does GPT 4.5 preview cost?

Input is priced at $75.00 per 1M tokens and output at $150 per 1M tokens (OpenAI, last verified May 7, 2026).

What is the context window of GPT 4.5 preview?

GPT 4.5 preview supports a 128,000-token context window with up to 16,384 output tokens.

Does GPT 4.5 preview support function calling?

Yes — GPT 4.5 preview supports function (tool) calling, including parallel tool calls.

Is GPT 4.5 preview good for production?

GPT 4.5 preview is well-suited for pricing — cheaper than 99% of priced chat models on Future AGI and PDF input — only 16% of chat models on Future AGI advertise this.

How can I route to GPT 4.5 preview with fallback?

Use Agent Command Center: a single OpenAI-compatible endpoint that supports cost-optimized routing, latency-aware retries, model fallback, and shadow traffic. Configure once, swap models without app changes.

Useful links for GPT 4.5 preview

Official sources, independent benchmarks, and pricing aggregators — no random search-engine guesses.