Cohere (tracing) | Respan Docs

Set up Respan

Sign up - Create an account at platform.respan.ai
Create an API key - Generate one on the API keys page

Use Respan Gateway

See Cohere gateway setup to route this integration through the Respan gateway.

Use AI

Add the Docs MCP to your AI coding tool to get help building with Respan. No API key needed.

1 {
2   "mcpServers": {
3     "respan-docs": {
4       "url": "https://mcp.respan.ai/mcp/docs"
5     }
6   }
7 }

What is Cohere?

The Cohere Python SDK provides chat, streaming chat, embedding, and rerank APIs. respan-instrumentation-cohere activates the Cohere OpenTelemetry wrapper and normalizes the emitted spans to the Respan span contract before export.

Setup

Install packages

$ pip install respan-ai respan-instrumentation-cohere cohere python-dotenv

Set environment variables

$ export RESPAN_API_KEY="YOUR_RESPAN_API_KEY"
$ export CO_API_KEY="YOUR_COHERE_API_KEY"

Optional:

$ export RESPAN_BASE_URL="https://api.respan.ai/api"
$ export COHERE_CHAT_MODEL="command-a-03-2025"

Initialize and run

1 import os
2 
3 import cohere
4 from dotenv import load_dotenv
5 from respan import Respan, workflow
6 from respan_instrumentation_cohere import CohereInstrumentor
7 
8 load_dotenv()
9 
10 respan = Respan(
11     api_key=os.environ["RESPAN_API_KEY"],
12     base_url=os.getenv("RESPAN_BASE_URL", "https://api.respan.ai/api"),
13     instrumentations=[CohereInstrumentor()],
14 )
15 client = cohere.ClientV2(api_key=os.environ["CO_API_KEY"])
16 
17 
18 @workflow(name="cohere_chat.workflow")
19 def run_chat() -> str:
20     response = client.chat(
21         model=os.getenv("COHERE_CHAT_MODEL", "command-a-03-2025"),
22         messages=[
23             {
24                 "role": "user",
25                 "content": "In one sentence, explain what Cohere rerank is useful for.",
26             }
27         ],
28         temperature=0.1,
29         max_tokens=80,
30     )
31     return response.message.content[0].text
32 
33 
34 try:
35     with respan.propagate_attributes(
36         trace_group_identifier="cohere_chat.workflow",
37         metadata={"workflow_name": "cohere_chat.workflow"},
38     ):
39         print(run_chat())
40 finally:
41     respan.telemetry.flush()

View your trace

Open the Traces page and search for the workflow name cohere_chat.workflow.

Configuration

Parameter	Type	Default	Description
`exception_logger`	`logging.Logger \| None`	`None`	Optional logger passed to the underlying Cohere OpenTelemetry instrumentor.
`use_legacy_attributes`	`bool`	`True`	Emits span attributes compatible with the existing Respan span contract.
`**instrumentor_kwargs`	`dict`	`{}`	Extra keyword arguments forwarded to the upstream instrumentor.

Attributes

Attach customer identifiers, thread IDs, and metadata to Cohere calls with propagate_attributes.

1 with respan.propagate_attributes(
2     customer_identifier="user_123",
3     thread_identifier="conversation_456",
4     trace_group_identifier="cohere_support_chat.workflow",
5     metadata={"plan": "pro", "workflow_name": "cohere_support_chat.workflow"},
6 ):
7     response = client.chat(
8         model="command-a-03-2025",
9         messages=[{"role": "user", "content": "Summarize our support policy."}],
10     )

Attribute	Type	Description
`customer_identifier`	`str`	Identifies the end user in Respan analytics.
`thread_identifier`	`str`	Groups related messages into a conversation.
`trace_group_identifier`	`str`	Groups spans by workflow name.
`metadata`	`dict`	Custom key-value pairs merged with default metadata.

Examples

Streaming chat

1 stream = client.chat_stream(
2     model="command-a-03-2025",
3     messages=[{"role": "user", "content": "Write a short haiku about observability."}],
4 )
5 
6 for event in stream:
7     if event.type == "content-delta":
8         print(event.delta.message.content.text, end="", flush=True)

Embeddings

1 response = client.embed(
2     model="embed-v4.0",
3     input_type="search_document",
4     texts=[
5         "Respan records traces and spans for AI systems.",
6         "Cohere embeddings convert text into vectors.",
7     ],
8     embedding_types=["float"],
9 )
10 print(len(response.embeddings.float_))

Rerank

1 response = client.rerank(
2     model="rerank-v3.5",
3     query="Which document explains search result ranking?",
4     documents=[
5         "Rerank models reorder candidate documents by query relevance.",
6         "Chat models generate conversational responses.",
7     ],
8     top_n=1,
9 )
10 print(response.results[0].index)