IBM Granite 4 H Small

IBM watsonx chat

IBM Granite 4 H Small is an IBM watsonx chat model.It supports a 20,480-token context windowwith up to 20,480 output tokens.Input is priced at $0.0600/M tokens and output at $0.250/M tokens. Capabilities include function calling. Route IBM Granite 4 H Small 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 12, 2026 View source ↗
Cost calculator

Estimate IBM Granite 4 H Small spend

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

~3K in / ~400 out · 5K req/day
3,000
020,480
400
020,480
5,000
01,000,000
Per request
$0.000280
in $0.000180 · out $0.000100
Per day
$1.40
5,000 requests
Per month
$42.61
152,188 requests

Estimate uses $0.0600/M input · $0.2500/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.

Pricing

Per-token rates, expressed in USD per 1M tokens. Verified May 12, 2026.

Input $0.0600/M
Output $0.250/M

Limits

Context window
20,480 tokens
Max input
20,480 tokens
Max output
20,480 tokens
Modalities
text

Capabilities

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

Where it's strong

  • +parallel tool calls — only 21% of chat models on Future AGI advertise this

Watch out for

  • !high cost — input + output rates are in the top 89% of priced chat peers; consider a cheaper sibling for high-volume workloads
  • !limited context — 20,480-token window is in the bottom quartile; not ideal for long documents or large RAG
  • !strict structured output — no JSON-schema enforcement, expect retry loops

Benchmarks pending

We haven't logged public benchmark scores for IBM Granite 4 H Small yet. Have one to contribute? Submit a source — citations help us prioritise.

Try it

Call IBM Granite 4 H Small 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.
# IBM Granite 4 H Small 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="watsonx/ibm-granite-4-h-small",
    messages=[{"role": "user", "content": "Hello, IBM Granite 4 H Small!"}],
)

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="watsonx/ibm-granite-4-h-small",
    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 ↗

Compare with similar models

IBM Granite 4 H Small doesn't have a public Arena ELO score yet, so we group by provider only — quality-tier comparisons need a benchmark.

FAQ

How much does IBM Granite 4 H Small cost?

Input is priced at $0.0600 per 1M tokens and output at $0.250 per 1M tokens (IBM watsonx, last verified May 12, 2026).

What is the context window of IBM Granite 4 H Small?

IBM Granite 4 H Small supports a 20,480-token context window with up to 20,480 output tokens.

Does IBM Granite 4 H Small support function calling?

Yes — IBM Granite 4 H Small supports function (tool) calling, including parallel tool calls.

Is IBM Granite 4 H Small good for production?

IBM Granite 4 H Small is well-suited for parallel tool calls — only 21% of chat models on Future AGI advertise this. Consider alternatives if you need high cost — input + output rates are in the top 89% of priced chat peers; consider a cheaper sibling for high-volume workloads.

How can I route to IBM Granite 4 H Small 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 IBM Granite 4 H Small

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